mechanical and viscoelastic properties of cementitious
TRANSCRIPT
Department of Civil Engineering
Mechanical and Viscoelastic Properties of Cementitious
Materials using Statistical Analysis Tools and
Nanoindentation
Hyuk Lee
This thesis is presented for the Degree of
Doctor of Philosophy
of
Curtin University
August 2016
Abstract
The research presents an investigation of mechanical and nanostructure characteristics of alter-native cementitious materials such as blended and alkali-activated cements using nanoinden-tation technology and statistical analysis tools. Statistical analysis of these mechanical andnano characteristics of cementitious materials results in materials properties which are useful instructural engineering application. The indentation analysis shows that indentation propertiessuch as modulus, hardness which are related respectively to elastic and strength properties withthe projected area during the indentation loading and unloading process. With microporome-chanics, the indentation modulus and hardness represents particle properties such as stiffness,Poisson’s ratio, cohesion, friction coefficient and packing density. The indentation analysis ofcementitious material identifies the link between the multiphase compositions, and elastic, mi-crostructure and viscoelastic properties. Statistical analysis tools are successfully being appliedas a useful tool in studying the influence of parameters in cementitious materials. The resultscan be analysed using the ANOVA technique to examine the variation in the measured prop-erties of cementitious materials. Moreover, the potential impact of indentation approach willencourage consideration of small scales examination to represent the large scale testing of civilengineering structural elements. Properties of hydration products of ordinary Portland cement(OPC) are determined as indentation modulus, hardness and packing density. The indentationmodulus, hardness and packing density of LD CSH are 16.787 ± 4.804 GPa, 0.704 ± 0.144GPa and 0.556 ± 0.01, respectively. The HD CSH has indentation modulus = 30.481 ± 4.257GPa, hardness = 1.415 ± 0.222 GPa and packing density = 0.595 ± 0.001. Creep compliancerate of OPC shows that capillary pores phase is the main phase that tends to increase creepcompliance. Also, indentation stress-strain curve shows that the strength failure can occur incapillary pore phase. Thus, porosity is an important factor to be considered when design OPCmixture. Based on statistic analysis of blended cements, an increase in fly ash content andsand to cementitious material ratio decreased the compressive strength development. The op-timization of the compressive strength of blended cement mixture is found to be 20% of fly ashcontent, 1.5 of sand to cementitious material ratio, 0.35 of water to cementitious material ratioand 0.2% of superplasticiser. Similarly, alkali-activated fly ash cement (AAFA) was investigatedbased on statistical analysis tools and nanoindentation. The results show that the increase insand has the greatest contribution to the increase in density. For compressive strength, normalpaste without SF, sand and SP with l/s of 0.6 gives the highest strength and the increase inSF significantly contributes to the adverse effect on compressive strength. For the indentationdata, the analysis using deconvolution technique confirms the four phases of reaction productsof AAFA. The main phase is sodium aluminosilicate hydrate (N-A-S-H), which is over 40%of the volume fraction. The microporomechanics of AAFA paste and mortar also demonstratethe relationships between the N-A-S-H volume fraction and strength; and activation degree andstrength. The creep behaviour of AAFA study revealed that partly-activated and non-activatedphases are the main reason for creep due to “block-polymerisation” concept. It is also foundthat liquid to solid ratio is the most affecting parameter on creep, an increase of liquid to solidratio leads to more creep.
ii
Acknowledgements
Firstly, my deepest gratitude goes to my supervisor, Dr. Vanissorn Vimonsatit, for hercontinuous guidance, encouragement and support during the pursue of my doctoral degree atCurtin University.
My gratitude also goes to my Research Guide Professor Prinya Chindaprasirt, Sustainabil-ity Infrastructure Development and Research Centre, Department of Civil Engineering, Facultyof Engineering, Khon Kaen University, Thailand, for providing a platform for completing thisresearch. I have a deep sense of gratitude and feel proud to have him as guide.
I would also like to acknowledge the assistance by Dr. Kornkanok Boonserm, Departmentof Applied Chemistry, Rajamangala University of Technology in Thailand, for her help withmicrostructural properties testing.
I would like to acknowledge the assistance and help this research by final year students HyoilKim and Yuyan Yang. I also thank Mick Elliss, Ashley Hughes, Dr Arne Bredin, Luke English,and Craig Gwyther, and others for their help with laboratory work at the Department of CivilEngineering, and special thanks to my friend, PengLoy Chow, who helped me to prepare in-dentation process.
I also want to express my special thanks to my father and mother for their continuous supporttill the completion of this Thesis. Lastly, this thesis would not be completed without the loveand support from my wife and son.
iii
Contents
1 Introduction 21.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2 Research Aim and Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.3 Research Significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.4 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Literature Reviews 62.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2 Construction Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Ordinary Portland cement . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2.2 Fly ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.2.3 Lime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Alkali-Activated Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.4 Alkali-Activated Lime-Pozzolan Cement . . . . . . . . . . . . . . . . . . . . . . . 142.5 Characteristic of Binding Cementitious Products . . . . . . . . . . . . . . . . . . 162.6 Engineering Statistics Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182.7 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Nanoindentation 213.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213.2 Homogeneous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.1 Types of Indenter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.2.2 Indentation Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Heterogeneous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253.4 Microporomechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273.5 Deconvolution Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.6 Indentation Fracture Toughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.7 Finite Element Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.8 Time-depending Nanoindentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.8.1 Indentation Contact Relaxation Modulus . . . . . . . . . . . . . . . . . . 363.8.2 Indentation Contact Creep Compliance . . . . . . . . . . . . . . . . . . . 373.8.3 Numerical Analysis of Viscoelasticity . . . . . . . . . . . . . . . . . . . . . 39
3.9 Indentation Stress-Strain Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423.10 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4 Properties of Cement Materials 474.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474.2 Indentation Properties of Ordinary Portland cement . . . . . . . . . . . . . . . . 48
4.2.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.2.2 Statistical Indentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494.2.3 Viscoelastic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534.2.4 Indentation Stress-Strain Curve and Fracture Toughness . . . . . . . . . . 554.2.5 Relationship between Indentation Properties . . . . . . . . . . . . . . . . 58
4.3 Statistical Analysis of Properties of Blended Cement . . . . . . . . . . . . . . . . 614.3.1 Experimental Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.3.2 Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634.3.3 Compressive Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
iv
4.3.4 Water Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734.3.5 Regression Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754.3.6 High Temperature Exposure . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.4 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5 Properties of Alkali-Activated Cement 865.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865.2 Taguchi’s Design of Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875.3 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895.4 Indentation Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.4.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945.4.2 Statistical Indentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.5 Viscoelastic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085.5.1 Contact Relaxation Modulus . . . . . . . . . . . . . . . . . . . . . . . . . 1085.5.2 Contact Creep Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . 1115.5.3 Indentation Stress-Strain Curve and Fracture Toughness . . . . . . . . . . 1185.5.4 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
6 Strain-hardening Behaviour of Cementitious Composite 1226.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1226.2 Design of Strain-hardening Behaviour . . . . . . . . . . . . . . . . . . . . . . . . 1246.3 Single Fibre Pull-out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.3.1 Numerical Model and Validation . . . . . . . . . . . . . . . . . . . . . . . 1286.3.2 Taguchi’s Design of Experimental . . . . . . . . . . . . . . . . . . . . . . . 1326.3.3 Effect of Parameters on Maximum Pull-out Force . . . . . . . . . . . . . . 1336.3.4 Single Fibre Pull-out Test with Polyvinyl Alcohol (PVA) Fibre . . . . . . 135
6.4 High-Performance Fibre Reinforced Cementitious Composite . . . . . . . . . . . 1376.4.1 Compressive Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.5 Flexural Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1456.6 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
7 Conclusion and Future Research 1567.1 Summary of Main Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1567.2 Recommendation for Future Research . . . . . . . . . . . . . . . . . . . . . . . . 158
Appendices 159Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
References 214
v
List of Figures
2.1 Powers-Brownyard model of two types of porosity (Powers, 1958) . . . . . . . . . 82.2 Descriptive model for alkali activation of aluminosilicate (Shi et al., 2011) . . . . 122.3 Effect of Na2SO3 dosage on strength development of LPC paste (Shi and Day,
1993b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 Relationship between the Ultimate strength and curing temperature (Shi and
Day, 1993a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.5 Effect of Lime content on Strength development of Lime-pozzolan cement cured
at 50◦C (Shi and Day, 1993a) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.6 General process of design of experiment (Park, 2007) . . . . . . . . . . . . . . . . 18
3.1 Schematic of contact between a rigid indenter and a flat specimen with modulusE (adapted after (Fischer-Cripps and Mustafaev, 2000)) . . . . . . . . . . . . . . 22
3.2 Typical indentation load-displacement curve . . . . . . . . . . . . . . . . . . . . . 233.3 Different types of indenter (a) Spherical (b) Berkovich (c) Conical (d) Vickers
indenter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.4 Berkovich geometry of contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.5 Element volume of material in a microscopic structure: (a) heterogeneous (b)
local heterogeneous (c) homogeneous (adapted after (Dormieux et al., 2006)) . . 263.6 Schematic representation of grid indentation technique: (a) average composite
material with large indentation depth (h� D), (b) low indentation depth withone phase (h � D),(c) grid indentation technique giving several phases usinglow indentation depth (h� D) adapted after (Constantinides et al., 2006)) . . . 27
3.7 Berkovich indentation crack parameters (adopted after (Ling, 2011)) . . . . . . . 303.8 Indentation load (P ) and depth (h) curve represent energies (adapted after
(Cheng and Cheng, 2004)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.9 Indentation geometry probe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323.10 Viscoelastic behaviour (a) ideal, (b) actual viscous fluid . . . . . . . . . . . . . . 333.11 Relaxation and creep stress with time . . . . . . . . . . . . . . . . . . . . . . . . 343.12 Mechanical models of time depending properties of material (a) Maxwell model
(b) Kelvin-Voigt Model (c) Combined Maxwell-Kelvin-Voigt model . . . . . . . . 363.13 A conical indenter impress on specimen . . . . . . . . . . . . . . . . . . . . . . . 393.14 Meshing configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403.15 Normalised contact relaxation modulus with analytical, numerical solution and
Maxwell model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403.16 Numerical and Rheological (Maxwell) solution of contact creep compliance . . . . 413.17 Contact creep compliance rate with numerical analysis . . . . . . . . . . . . . . . 423.18 Schematic uniaxial and indentation stress-strain curve (adopted after (Martinez
et al., 2003)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433.19 Indentation stress-strain curve on Fused silica with Spherical tip . . . . . . . . . 443.20 Indentation stress-strain curve on Fused silica with Berkovich tip . . . . . . . . . 453.21 The deformation regimes with different types of indenter tip (adapted after (Mar-
tinez et al., 2003)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.1 Deconvolution technique of indentation modulus and hardness . . . . . . . . . . . 504.2 Typical indentation load-depth (P − h) curves . . . . . . . . . . . . . . . . . . . 504.3 Packing density (η) distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514.4 Deconvolution result of packing density . . . . . . . . . . . . . . . . . . . . . . . 51
vi
4.5 Contour map of hydration products of OPC paste. Image size is 180µm × 180µmwith 20µm grid spacing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.6 Normalised relaxation modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544.7 Typical logarithm curve fitting in creep phase . . . . . . . . . . . . . . . . . . . . 544.8 Contact creep compliance rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.9 Deconvolution result of contact creep modulus . . . . . . . . . . . . . . . . . . . 564.10 Contact creep compliance rate of hydration phases . . . . . . . . . . . . . . . . . 574.11 Indentation stress-strain curves (a) MP (b) LD CSH (c) HD CSH (d) CH . . . . 584.12 Deconvoluted indentation properties with Power fit . . . . . . . . . . . . . . . . . 604.13 XRD pattern of (a) OPC (Type I) and (b) low calcium fly ash . . . . . . . . . . 624.14 Density with curing age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634.15 Density with curing age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644.16 Contribution of experimental parameters on density . . . . . . . . . . . . . . . . 654.17 (a) compressive strength development and (b) % of compressive strength devel-
opment between 7 to 28 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674.18 XRD pattern of Mix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684.19 SEM image of Mix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684.20 Effect of parameters on compressive strength . . . . . . . . . . . . . . . . . . . . 704.21 Contribution of experimental parameters on compressive strength . . . . . . . . . 714.22 Effect of parameters on compressive strength gain . . . . . . . . . . . . . . . . . 724.23 Contribution of experimental parameters on compressive strength gain . . . . . . 734.24 Water absorption results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734.25 Effect of parameters on water absorption results . . . . . . . . . . . . . . . . . . 744.26 Contribution of experimental parameters on water absorption . . . . . . . . . . . 754.27 Residual strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784.28 Residual strength reduction factors (adopted after (de Normalisation, 2005)) . . 794.29 XRD pattern of Mix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804.30 SEM image of Mix 1 after exposed to high temperatures . . . . . . . . . . . . . . 804.31 Effect of parameters after exposed to high temperatures . . . . . . . . . . . . . . 814.32 Predicted residual strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.1 X-ray diffraction pattern of low calcium fly ash . . . . . . . . . . . . . . . . . . . 885.2 Effect of parameters on density and compressive strength at 28 days . . . . . . . 905.3 Contribution of experimental parameters on (a) density (b) compressive strength 915.4 Typical indentation load-depth (P − h) curve on AAFA . . . . . . . . . . . . . . 955.5 Packing density relationship distribution of AAFA on Mix 1 . . . . . . . . . . . . 955.6 Effect of parameters on reaction properties (a) Stiffness (b) Cohesion (c) Friction
angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975.7 Effect of parameters on reaction properties (a) Stiffness (b) Cohesion (c) Friction
angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985.8 Contribution of experimental parameters on N-A-S-H phase at 28 day . . . . . . 1025.9 Contribution of experimental parameters on activated degree and porosity . . . . 1045.10 Volume fraction distribution with varying silica fume of (a) reaction products
(b) porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055.11 Volume fraction distribution with varying liquid to solid ratio of (a) reaction
products (b) porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1065.12 Relationship between compressive strength and volume fraction of N-A-S-H phase1075.13 Relationship between compressive strength and degree of activation . . . . . . . . 1075.14 Normalised relaxation modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1105.15 Contact creep compliance rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125.16 Effect of parameters on creep modulus . . . . . . . . . . . . . . . . . . . . . . . . 1135.17 Contribution of experimental parameters on creep modulus . . . . . . . . . . . . 1135.18 Effect of parameters on creep modulus of partly-activated and non-activated phases1165.19 Contribution of experimental parameters on creep modulus of (a) Partly-activated
slag (b) Non-activated slag (c) Non-activated compact glass . . . . . . . . . . . . 1175.20 Indentation stress-strain curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1185.21 Effect of parameters on fracture toughness . . . . . . . . . . . . . . . . . . . . . . 1195.22 Contribution of experimental parameters on fracture toughness . . . . . . . . . . 120
vii
6.1 The concept of stress-strain hardening and strain softening under tensile stress . 1236.2 HPFRCC stress-strain behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236.3 Tensile stress-strain curve for HPFRCC (adopted after (Kanda et al., 2000)) . . . 1266.4 Idealised interface law in three stages for single fibre pull-out (adopted after
(Zhan and Meschke, 2014)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1286.5 Fracture Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1296.6 Single fibre pull-out simulation model without inclined angle . . . . . . . . . . . 1296.7 Meshing configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1296.8 Equivalent stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1306.9 Validation of FE model with analytical model . . . . . . . . . . . . . . . . . . . 1316.10 S/N ratio of single fibre pull-out . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336.11 Parameter contribution on fibre pull-out test . . . . . . . . . . . . . . . . . . . . 1346.12 Schematic of single fibre pull-out test . . . . . . . . . . . . . . . . . . . . . . . . . 1356.13 XRD pattern of fly ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376.14 A schematic of flexural performance . . . . . . . . . . . . . . . . . . . . . . . . . 1386.15 Compressive strength development of AAFA composites,Series A . . . . . . . . . 1396.16 Compressive strength development of OPC composites, Series P . . . . . . . . . 1406.17 Typical stress-strain curves of AAFA composites, Series A . . . . . . . . . . . . . 1416.18 Typical stress-strain curve of OPC composites, Series P . . . . . . . . . . . . . . 1426.19 The effect of fibre volume ratio on strain and compressive strength, Series A . . . 1436.20 Flexural behaviour of AAFA composites in Group A of Series A . . . . . . . . . . 1466.21 Flexural behaviour of AAFA composites in Group B of Series A . . . . . . . . . . 1466.22 Flexural behaviour of OPC composite, Series P . . . . . . . . . . . . . . . . . . 1476.23 Effect of fibre volume fraction on deflection capacity in AAFA composites, Series A1486.24 Effect of fibre volume fraction on flexural strength in AAFA composites, Series A 1496.25 Effect of fibre volume fraction on toughness in AAFA composites, Series A . . . . 1496.26 Effect of fibre volume fraction on deflection of OPC composites, Series P . . . . . 1506.27 Effect of fibre volume fraction on flexural strength of OPC composites, Series P . 1516.28 Effect of fibre volume fraction on toughness of OPC composites, Series P . . . . 1516.29 Critical volume fraction against interfacial bond strength with AAFA composites,
Series A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1546.30 Critical volume fraction against interfacial bond strength with OPC composites,
Series P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
viii
List of Tables
2.1 General features of types of cement in ASTM C150 (Neville, 2011) . . . . . . . . 72.2 AS 3582: Grade Specified requirement (Australia Standard, 2009) . . . . . . . . 92.3 ASTM C618 : Chemical Requirements of fly ash (ASTM Standard C618, 2012) . 92.4 Major element chemistry of fly ash in selected countries (adopted (French and
Smitham, 2007)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.5 Application of Geopolymer material based on silica to alumina atomic ratio (Ran-
gan, 2008) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132.6 Terminologies of design of experiment . . . . . . . . . . . . . . . . . . . . . . . . 18
3.1 Project contact area, geometry factors and intercept factor for various types ofindenters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Different indentation area function from numerical analysis (Sakharova et al.,2009) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3 Indentation geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.1 General features of types of cement in ASTM C150 (Neville, 2011) . . . . . . . . 474.2 Properties and characteristics of cements in AS 3972-2010 (Australia Standard,
2010) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.3 Deconvolution results for indentation modulus and hardness . . . . . . . . . . . . 494.4 Deconvolution results for indentation modulus and hardness . . . . . . . . . . . . 514.5 Rheological properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534.6 Deconvolution results of indentation modulus, hardness, contact creep modulus
and packing density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.7 Deconvolution results of indentation modulus, hardness, contact creep modulus
and packing density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.8 Variation parameters and levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.9 Standard L9 orthogonal array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.10 Chemical composition of OPC (Type I) and low calcium fly ash (wt. %) . . . . . 624.11 Density with curing age results (kg/m3) . . . . . . . . . . . . . . . . . . . . . . . 644.12 ANOVA results on density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644.13 The results in based on compressive strength (MPa) . . . . . . . . . . . . . . . . 664.14 ANOVA results on compressive strength . . . . . . . . . . . . . . . . . . . . . . . 694.15 ANOVA results on density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724.16 ANOVA results on density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744.17 Experiment and Predicted density (kg/m3) and compressive strength (MPa) at
28 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764.18 Regression analysis of test data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764.19 Residual strength of blended cement mixtures (Unit in MPa) . . . . . . . . . . . 784.20 ANOVA results on high temperatures exposure . . . . . . . . . . . . . . . . . . . 814.21 Predicted S/N ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824.22 Coefficient of empirical relationship in Equation (4.3.3) . . . . . . . . . . . . . . . 824.23 Regression analysis of residual compressive strength . . . . . . . . . . . . . . . . 83
5.1 Variation parameters and levels for AAFA mixture . . . . . . . . . . . . . . . . . 875.2 Standard L9 orthogonal array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875.3 Chemical composition (wt.%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885.4 Mix proportion with density and compressive strength test result . . . . . . . . . 895.5 ANOVA on density and compressive strength at 28 days . . . . . . . . . . . . . . 89
ix
5.6 Experiment and Predicted density and compressive strength at 28 days . . . . . 925.7 Regression analysis of test data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935.8 Properties of the reaction products . . . . . . . . . . . . . . . . . . . . . . . . . . 955.9 ANOVA results on properties of reaction products . . . . . . . . . . . . . . . . . 965.10 Deconvolution results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005.11 Deconvolution results of Volume fraction (f) (Unit in %) . . . . . . . . . . . . . 1015.12 ANOVA results on properties of reaction products . . . . . . . . . . . . . . . . . 1015.13 ANOVA results on properties of reaction products . . . . . . . . . . . . . . . . . 1035.14 Degree of activation and porosity (Unit in %) . . . . . . . . . . . . . . . . . . . . 1045.15 Rheological properties (Unit in GPa) . . . . . . . . . . . . . . . . . . . . . . . . . 1085.16 Average logarithmic coefficients and corrected coefficient . . . . . . . . . . . . . . 1115.17 ANOVA results on creep modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135.18 Specific creep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145.19 Deconvolution results of Creep Modulus (C) (Unit in GPa) . . . . . . . . . . . . 1155.20 ANOVA results on creep modulus of partly-activated and non-activated phases . 1155.21 Fracture energy release rate and toughness . . . . . . . . . . . . . . . . . . . . . . 1195.22 ANOVA results on fracture toughness . . . . . . . . . . . . . . . . . . . . . . . . 120
6.1 Variation parameters and levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326.2 Standard L27 orthogonal array . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326.3 Numerical studies of single fibre pull-out with Taguchi’s DOE . . . . . . . . . . . 1336.4 ANOVA of fibre pull-out force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1346.5 Maximum interfacial bonding strength . . . . . . . . . . . . . . . . . . . . . . . . 1366.6 Parameters for analytical modelling of fibre pull-out . . . . . . . . . . . . . . . . 1366.7 Maximum pull-out force between analytical and experimental results . . . . . . . 1366.8 Chemical composition of OPC (type I) and low calcium fly ash (wt. %) . . . . . 1376.9 Composition of mix proportions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386.10 Compressive strength development (MPa) . . . . . . . . . . . . . . . . . . . . . . 1406.11 Compressive strain and strength test results . . . . . . . . . . . . . . . . . . . . . 1426.12 Modulus of Elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1446.13 Flexural behaviours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
A1 Indentation test of OPC paste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159B1 Indentation test of AAFA mix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163B2 Indentation test of AAFA mix 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168B3 Indentation test of AAFA mix 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172B4 Indentation test of AAFA mix 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176B5 Indentation test of AAFA mix 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181B6 Indentation test of AAFA mix 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185B7 Indentation test of AAFA mix 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190B8 Indentation test of AAFA mix 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194B9 Indentation test of AAFA mix 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
1
Chapter 1
Introduction
1.1 General
In the last decades, consumption of construction materials such as concrete and mortar in the
world is highly increasing due to population and economy growth, which results in high de-
mand for buildings, roads, and infrastructures. Concrete is the most widely used construction
material because of its low cost, ease of placement, flexibility to form into shapes, and avail-
ability locally. Concrete requires cementitious materials that bind the construction materials
together. However, traditional cementitious materials such as Portland cement are now facing
environmental issues due to carbon dioxide emission during the cement production. Carbon
dioxide emission is mainly a result of limestone decarbonation which is necessary in the chem-
istry during the cement production (Flatt et al., 2012). It is estimated that cement production
is releasing approximately 5 to 8 percentage of the total carbon dioxide emission per year, and
is predicted to reach a rate of 3.5 billion tonnes per year by 2025 (Van den Heede and De Belie,
2012). Therefore, development of alternative cementitious materials to reduce the environmen-
tal impact caused by high carbon dioxide emission from cement production a major subject of
on-going intensive research worldwide.
A type of cementitious materials can be developed by using pozzolan as a partial replacement of
Portland cement (Neville, 2011). This type of cements is generally called blended cements, and
the common types of pozzolan used in blended cements are industry by-products such as fly ash
and ground granulated iron blast-furnace slag. Application of blend cements has been studied
particularly on the improved mechanical properties of concrete made with blended cements.
Despite the intensive research on different applications of blended cements, slow setting time
and early strength development due to poor pozzolans reaction at early age are negative effects
of using this type of cements. These negative effects thus limit the content of pozzolans to be
used. However, blended cements are still preferred with the degree of cement replacement as
high as 35 percent, depending on the properties of available pozzolans.
Another type of alternative cements is alkali-activated cement, which is made mainly of poz-
zolans and uses alkali activators to form synthetic binder product. Research on alkali-activated
cements for application of construction materials are currently under ongoing development.
Suitable pozzolans for alkali-activated cements must possess high quantities of reactive silicate
and aluminate compounds. Fly ash, ground granulated iron blast-furnace slag, and metakaolin
are most commonly used as alumina silicate sources. The main positive effect of alkali-activated
cements is that the source materials are not carbon bearing materials, thus, do not release vast
quantities of carbon dioxide compared to traditional cementitious materials such as Portland
2
cement. Other important advantages of alkali-activated cements are their superior mechanical,
chemical, and thermal properties compared to traditional cementitious materials. In spite of
advantages of alkali-activated cements, application of alkali-activated cements is still limited
mainly due to quality control of source materials and sensitivity to alkali concentration. These
disadvantages discourage their application in construction.
Several earlier studies have been focussed on development of blended cements and alkali-
activated cements as alternative cementitious materials. Notwithstanding a wide range of
investigation of these alternative cementitious materials, there is still a lack of knowledge of
properties in micro and nanoscale and the interaction between constituents during the binding
process, which govern mechanical properties such as strength, durability, and ductility. Tradi-
tional design of cementitious materials is aimed at the final properties of the binded products
which do not sufficiently provide fundamental knowledge of the binding properties. Recently,
however, industries have realised the need to evaluate the fundamental properties of cemen-
titious materials for manipulating materials to meet the needs of modern technology. A way
to establish a knowledge in evaluating properties of alternative cementitious materials is by
exploring the possibility of improving the knowledge mathematically. This method necessi-
tates selecting appropriate statistical analysis tools for evaluating the knowledge of properties.
These statistical analysis tools can provide a meaningful measurement of properties. Advance
in these cementitious materials, therefore, can be preceded almost every major technological
leap in civil engineering. This need, in turn, is driving research to develop and investigate
fundamental properties of cementitious materials. This research is to investigate fundamen-
tal and interaction properties between constituents of binding process in blended cements and
alkali-activated cements. The outcome will enhance a development of cementitious materials
and increase their potential applications in civil engineering and construction.
1.2 Research Aim and Objective
This research is aimed at investigating properties of cementitious materials to determine me-
chanical property, chemical property, microstructure, and nanostructure of blended cements
and alkali-activated fly ash based cements based on experiment and statistical analysis of test
data.
To achieve the aim, the objectives are identified in 4 parts, which are to determine various
properties of different cementitious materials as follows:
- Part 1: Nanostructure properties of cementitious materials:
◦ To assess the reaction products of blended cements and alkali-activated fly ash based
cements, and by using nanotechnology and statistical analysis tools to evaluate inter-
action between constituents binding process and govern mechanical properties such
as modulus, hardness, packing density and microporomechanics of the materials;
◦ To apply microporomechanics on the tested materials to identify interactive proper-
ties of the materials;
◦ To investigate nano properties of viscoelasticity such as relaxation and creep in the
tested cementitious materials based on short-term test, and to use statistical analy-
sis tools to use statistical analysis methods to determine long-term time-dependent
behaviour of the materials.
- Part 2: Properties of Portland cement and blended cements
3
◦ To determine nanostructure characteristics of Portland cement, which is traditional
cementitious material using nanotechnology.
◦ To determine properties of blended cements; these properties are density, compressive
strength, high temperature resistance and water absorption, and to use statistical
analysis tools to evaluate interactive relationships between constituent ingredients.
- Part 3: Properties of alkali-activated fly ash based cementitious systems
◦ To investigate properties of alkali-activated fly ash based cement, which are density
and compressive strength, using statistical analysis tools; and to identify the effect
of constituent ingredients on the alkali-activated fly ash based cementitious systems.
◦ To determine characteristics of alkali-activated fly ash based cement using nanotech-
nology and statistical analysis tools, and to evaluate the relationship between nanos-
tructure and overall alkali-activated fly ash based cementitious systems.
- Part 4: Analysis of properties of cementitious materials for use in structural engineering
application.
◦ To investigate the effect of cementitious properties on strain-hardening behaviour of
composite; and to examine the structural performance of composites with respect to
flexure and compressive strengths;
◦ To evaluate influence of nano characteristic on structural performance of composites.
1.3 Research Significance
This research contributes to the area of research relating to the development of sustainable,
alternative cements and nano characteristics of cementitious materials. The investigation of
nanostructure characteristics of cementitious materials will be a significant part of this research.
The outcome will pose numerous pertinent questions to guide future research. The main points
of significance are as follows:
- Contribute to evaluation of the fundamental properties of cementitious materials for ma-
nipulating materials to meet the needs of modern technology.
- Enhance the development of constituent ingredients of cementitious material systems by
statistical expression.
- Contribute the development of alternative cementitious materials.
- Recommend strategies and guidelines design of cementitious materials to achieve strain-
hardening composites.
1.4 Outline of Thesis
The thesis contains the following Chapters. Chapter 2 reviews the literature on background of
cementitious systems, statistical analysis tools for engineering, and experimental methodologies
for determining chemical and microstructure properties. In Chapter 3, a review of theory of
nanotechnology and developed statistical analytical tools is provided. Nanotechnology theory
is described based on nanoindentation method which is developed from contact mechanics.
The major challenge in this Chapter is applying nanotechnology to characterise cementitious
materials. In particular, this Chapter presents a development of analytical tools for determin-
ing viscoelastic properties which are relevant in civil engineering application. Chapter 3 also
provides a theoretical framework for solving contact mechanics problems of nanoindentation.
4
Chapter 4 contains two main parts. The first part focuses on the nanostructure characteris-
tics of Portland cement systems. Nano characteristics are determined through the developed
nanotechnology applications. The second part of Chapter 4 focuses on a series of experiment
on blended cements and analysis of test data using statistical tools. The results lead to un-
cover interactive relationship between constituent ingredients of blended cements. Chapter 5
deals with mechanical and nano characteristics of alkali-activated fly ash based cements by
using statistical tools to determine the interactive relationship between mechanical and nano
characteristics between constituent ingredients of alkali-activated fly ash based cements. Nan-
otechnology and microporomechanics are applied to determine nano characteristics, which lead
to meaningful information of reaction properties of the materials. Chapter 6 presents a study
on the application of the results of mechanical and nano characteristics of cementitious materi-
als in determining strain-hardening behaviours of the materials. A design guideline to achieve
strain-hardening composites is provided. Finally, Chapter 7 summarises the outcomes of this
research and provides suggestions for further research.
5
Chapter 2
Literature Reviews
2.1 Introduction
This Chapter presents a detailed background on cementitious materials. A considerable amount
of research has been undertaken into determining the properties of alkali-activated cement in
respect to constituents, activators, reaction products, and microstructures. A critical review of
findings reported in the literature forms a fundamental knowledge for this research.
2.2 Construction Materials
2.2.1 Ordinary Portland cement
Ordinary Portland cement (OPC) is made primarily from calcareous materials such as limestone
or chalk, and procedure alumina and silica found as clay or shale. The manufacturing process of
OPC consists of grinding raw materials and mixing them intimately in certain proportions and
burning at a temperature up to about 1450◦C in a rotary kiln. During heating, partial fusion
happens, and nodules of clinker are produced. The clinker is then mixed with few percentage
of gypsum which is to control the setting rate of binder product. Grinding and mixing of raw
materials can be done either in a wet or dry condition (Neville, 2011). Alite (C3S), Belite(C2),
Aluminate (C3A) and Ferrite (C4AF) are generally the major constituents of OPC: Common
chemical terms and their notations are: CaO = C, SiO2 = S, Al2O3 = A and Fe2O3 = F. C3S
and C2S are the most important constituents of OPC, their contents are 50-70% and 15-30%
of total, respectively (Neville, 2011). These main constituents of OPC when mixed with water
will produce Calcium Silicate Hydrate (CSH), Portlandite (CH), AFm (Al2O3Fe2O3-mono)
and AFt (C3A·3CaO·SO3H12). ASTM ASTM Standard C150 (2015) has designed six type of
Portland cement as Type I to V and White cement, which differ primarily in C3A contents
and fineness. Table 2.1 shows the general characteristics of types of cement classified by ASTM
ASTM Standard C150 (2015). CSH is the main hydration product which has around 50 to 70%
of the volume of total hydration products, and governs fundamental properties of concrete such
as strength. Various experimental techniques such as mechanical, chemical and microstructural
techniques have been used to understand behaviour over the last 100 years. The description
of CSH was up to morphology and relationship between the microstructure and macroscopic
properties. According to Neville (2011), gel-like morphology CSH was observed by Le Chatelier
more than 100 years ago. Taylor (1997) found that CSH is amorphous and has the properties
of a rigid gel and its structure is similar to that of natural minerals such as Tobermorite
(Ca5Si6O16(OH)24H2O) and Jennite (Ca9Si6O18(OH)64H2O). Two types of semi-crystalline of
6
Table 2.1: General features of types of cement in ASTM C150 (Neville, 2011)
Type Classification Characteristics
I General Purpose High C3S contentII Sulphate resistance Low C3A content (< 8%)III High early strength Fine groundIV Low heat (slow reaction) Low C3S content (< 50%) and C3AV High sulphate resistance Low C3A content (< 5%)
White White colour Low MgO, No C4AF
CSH, CSH (I) and CSH (II) are reported (Taylor, 1997) to be a structurally imperfect form of
Tobermorite and Jennite, respectively. From the commotional perspective, Jennite and CSH
(II) are closer to CSH gel than Tobermorite and CSH (I) (Taylor, 1997).
The Powers-Brownyard model (Taylor, 1997) is one of the models of CSH, which provides
quantitative information on CSH properties. This model assumes that properties of the CSH
gel of cement are relatively rigid and strong - as solid of internal surface. Also, the model
described the broad structure of the material by a model based largely on an indication from
total and non-evaporable water contents and water vapour sorption isotherms. According to
Powers-Brownyard model (Powers, 1958), volumetric quantities of porosity of two phases such
as capillary porosity (VCP ), and gel porosity (VGP ) can be expressed by:
VCP = 0.20α
[1− w/c
w/c+ ρw/ρc
](2.2.1a)
VGP = 2.12α
[1− w/c
w/c+ ρw/ρc
](2.2.1b)
The following symbols are used in Equation (2.2.1) as:
w/c = Initial water to cement ratio
ρc = Density of cement, typically 3150kg/m3
ρw = Density of water, typically 1000kg/m3
α = The degree of hydration at w/c ≤ w/c∗
Based on the Powers-Brownyard model, the relationship between volumetric porosity of three
phases in Equation (2.2.1) can be graphically presented in shown in Figure 2.1. This figure shows
that an increase of water to cementitious ratio leads to a decrease of the volume fraction of two
types of porosity. The mechanical strength of OPC is required for engineering structures. The
strength of concrete or mortar mostly depends on the cohesion of OPC paste and its adhesion
to the aggregate particles. There are several forms of strength tests such as compression and
flexure. The compression test is a most common method to determine the strength for structural
design. Similarly, flexural strength generally gives a good knowledge of the strength in tension.
There are several empirical relationships between porosity and strength. The most satisfactory
empirical strength equation is as (Taylor, 1997):
σ = σo (1− Cη) (2.2.2)
where σo is hypothetical maximum compressive strength attainable, η is mean of total porosity
and C is constant.
7
Figure 2.1: Powers-Brownyard model of two types of porosity (Powers, 1958)
2.2.2 Fly ash
Coal combustion products (CCPs) are inorganic solid particulates that are produced from
burning coal to generate electricity. CCPs include fly ash, bottom ash, boiler slag, flue gas
desulfurization gypsum, and others by-products from power plants. Use of CCPs in identified
applications can have substantial environmental benefits and fly ashes represent 80 to 90 per-
centage of the total CCPs volume (Guide to the use of fly ash in concrete in Australia, 2009).
The beneficial uses of CCPs are:
- Waste stream reduction and associated reduction in the requirement for the landfill,
- Conservation of resources such as the use of gypsum, limestone or fly ash as a replacement
in cement product, and
- Reduction of greenhouse gas emission when used as a cement replacement.
Bottom ash is formed in pulverised coal furnaces and agglomerated ash particles. It is too
large to be passed in the flue gases, which exit to the atmosphere by flue and impinge on the
furnace wall, or fall through an open grate to an ash hopper at the bottom of the furnace.
Bottom ash is generally grey in black colour. It has porous surface structure with an angular
shape. Bottom ash can be used as a replacement of aggregate and is usually well graded in
size. The porous surface structure of bottom ash particles makes this material less durable than
conventional aggregates and better suited for use in the base course and shoulder mixtures or
cold mix applications, as opposed to wearing surface mixtures. This porous surface structure
is also suitable in lightweight concrete applications (U.S. Environmental Protection agency,
2007). Fly ash has been used in concrete in low cost and improves properties of concrete,
therefore, usage of fly ash is gradually increasing. The main contribution of fly ash is an
improvement in workability, mix efficiency and improved concrete placement characteristics
(Guide to the use of fly ash in concrete in Australia, 2009). Also, Shi (1996) explained that
one of the main contributions to the strength of fly ash is pozzolanic reaction between fly ash
and CH. Generally, fly ash is separated from filter bags and flue gases by mechanical collector
and electrostatic precipitators. The main difference between electrostatic precipitators and
filter bags is electrostatic classify particles size. As fly ash is a by-product material, carbon
8
emission is virtually none compared with OPC base cement. Only carbon emission due to the
transportation of fly ash is to be considered but it is still considerably low (Strategies, 2007;
Heidrich and Woodhead, 2005). There are two grades of fly ash which are classified in AS
3582 (Australia Standard, 2009) as Normal and Special grade. In some areas of Australia,
there are limited quantities available of highly reactive fly ash (Special grade). In some areas
of Australia, there are limited quantities of highly reactive fly ash (Special grade). Special
grade fly ash complies with the requirements of fine grade as shown in Table 2.2 the relative
compressive strength at 28 days of curing should exceed 105% compared to the strength of
normal grade. ASTM ASTM Standard C618 (2012) classifies fly ash, as Class C and Class F,
which is according to the total percentage of Silica dioxide (SiO2), Aluminium oxide (Al2O3)
and Iron oxide (Fe2O3) as shown in Table 2.3. Generally, Class C fly ash has calcium oxide
contents more than Class F fly ash. Sometimes, Class C fly ash and Class F fly are calling “High
calcium fly ash” and “Low-calcium fly ash”, respectively. Also, ASTM ASTM Standard C618
(2012) classifies that raw or calcined natural pozzolan such as opaline cherts, shales, tuff and
volcanic ashes or pumicites as Class N. French and Smitham (2007) presented major elements
of fly ash from different countries with significant variation in major elements chemistry. Table
2.4 shows major chemical compositions of fly ash derived from different counties SiO2, Al2O3,
and basic oxides are calcium oxide (CaO), Magnesium oxide (MgO), Iron (III) oxide (Fe2O3),
Sodium oxide (Na2O) and Potassium oxide (K2O) approximately. The main effect of fly ash in
cement system is that on water demand and workability. Neville (2011) noted that between 5
and 15% of fly ash is the reduction in water demand of concrete for a consistent workability.
The main reason for a decrease in water demand of concrete is ‘ball-bearing’ effect which is
ascribed to fly ash due to its spherical shape. At the same time, another mechanism is involved
that fine fly ash particles become adsorbed on the surface of the cement particle by electrical
Table 2.2: AS 3582: Grade Specified requirement (Australia Standard, 2009)
Grade Fine Medium Coarse
Fineness∗ (% minimum) 75 65 55Loss on ignition (% maximum) 4 5 5Moisture content (% maximum) 1 1 1
SO3 contents (% maximum) 3 3 3
∗ by mass passing 45µm sieve
Table 2.3: ASTM C618 : Chemical Requirements of fly ash (ASTM Standard C618, 2012)
Class F Class C
SiO2 + Al2O3 + Fe3O3 (% minimum) 70.0 50.0SiO3 (% maximum) 5.0 5.0
Moisture content (% maximum) 3.0 3.0Loss on ignition (% maximum) 6.0 6.0
Table 2.4: Major element chemistry of fly ash in selected countries (adopted (French and Smitham,2007))
Components Australia US UK Japan
SiO2 55 to 60% 55 to 63% 43 to 48% 55 to 62%Al2O3 24 to 30% 18 to 28% 25 to 30% 23 to 28%
Basic Oxides 15 to 28% 18 to 33% 25 to 30% 14 to 18%
9
charges. Thus, the water demand for a given workability is reduced because fine fly ash particles
are covering the surface of the cement particles. However, the excess amount of fly ash would be
no benefit on water demand. Approximately up to 20% of fly ash content is reported to reduce
water demand in concrete. According to Ramezanianpour (2014), Class C fly ash has self-
hardening properties such as C2S, C3A, CaSO4, MgO and free CaO. The hydration behaviour
of C2S and C3A in Class C fly ash is the same as that in OPC, but the hydration rate of CSH
form is comparatively slow. Somehow, thus, Class C fly ash when mixed with water produces
hydration products in OPC as AFm, AFt and CSH. Class F fly ash, however, has little or no
self-hardening properties. Class F fly ash only hydrates when alkalis and CH are added. The
hydration products such as CSH, C2ASH8 and C4AH are formed and they are produced in the
later stage of the hydration process. Generally, fly ash in concrete is to be incorporated because
packing action of the fly ash particles at the interface of coarse aggregate particles would result
in reduced permeability. However, the chemical reactions of fly ash have an effect in improving
the microstructure of hydrated cement pastes which can lead to the strength development of
concrete. Papadakis (1999) studied the effect of Class F fly ash on Portland cement system.
They found higher strength in concrete when using Class F fly ash to replace aggregates or
cement. When Class F fly ash reacts with CH; it gives higher water content and lower total
porosity (Papadakis, 1999). Poon et al. (2000) reported that high strength of concrete with
45% Class F fly ash has a lower heat of hydration and chloride diffusivity than normal plain
cement concrete. Class F fly ash in concrete with lower water to binder ratio has better strength
contribution. Another research from Papadakis (2000) reported that Class C fly ash content
in concrete directly led to the strength enhancement after the mixing. The reaction of Class C
fly ash in hydrating cement makes it lower porosity due to the high content of reactive calcium
bearing phase in fly ash.
2.2.3 Lime
There are two types of Lime, Quicklime and Hydrated Lime (slake lime) which are commonly
used in practice. According to Shi (1992), quicklime with cement has higher strength than
hydrated lime with cement when having the same content of quicklime and undergoing the same
curing condition. Quicklime with water react quickly and form reaction production of calcium
hydroxide which is very fresh and high reactivity than commercial hydrated lime. Quicklime
is commonly known in the form of Calcium Oxide (CaO) or burnt lime. It is produced from
limestone (CaCO3) by calcination at high temperature to decompose the limestone to quicklime.
It is an alkaline product with more than pH 12, and normally in powder or granule with a density
greater than 1000 kg/m3. The process of progressive change of limestone to lime is shown as:
CaCO3heat−−−→ CaO + CO2 (2.2.3)
Hydrated lime is the chemical content of Calcium hydroxide (Ca(OH)2). It is produced by
quicklime (CaO) mixing with water. It is an alkaline product with pH¿12 and fine power with
450 to 780 kg/m3 of density (What is Lime?, 2005).
CaO +H2O → Ca(OH)2 (2.2.4)
2.3 Alkali-Activated Cement
Purdon Pacheco-Torgal et al. (2008) introduced a major development of alkali-activated cement
in 1940s. A highly concentrated alkali hydroxide solution or silicate solution reacts with a solid
10
aluminosilicate produces a synthetic alkali aluminosilicate materials called “geopolymer” (Davi-
dovits, 1994b; Joseph, 2011). These materials are classified as polymer because their structures
are large molecules formed by a number of groups of smaller molecules. Theses polymer are
formed by the reaction of an alkali solution and a source material which is rich in aluminosilicate
and includes organic minerals such as kaolinite and inorganic materials such as fly ash (Kumar
et al., 2005). Depending on the source material selection and recession conditions, geopolymer
exhibits a variety of properties and characteristics such as strength, shrinkage, setting time,
acid resistance, fire resistance and thermal conductivity. Thus, it is not necessarily intrinsic
to formulation but correct mix and processing design to optimisation properties and reduce
the cost for application (Duxson, Fernandez-Jimenez, Provis, Lukey, Palomo and Van Deven-
ter, 2007). Generally, ‘geopolymer’ is also commonly referred to as ‘alkali-activated cement’
(Shi et al., 2011). In this research, these synthetic alkali aluminosilicate materials are called
‘alkali-activated cement’ (AAC). According to Shi et al. (2011), AAC mainly consists of two
components, which are cementitious components and alkali activators. Cementitious compo-
nents can be taken from a variety of industrial by-products composed of silica and aluminium
such as granulated blast furnace slag, volcanic glass, coal fly ash zeolite, metakaolin, silica fume
and nonferrous slag (Shi et al., 2011). The composition of the cementitious components and
alkali-activated cement can be classified into five categories:
- Alkali-activated slag-base cement
- Alkali-activated Pozzolan cement
- Alkali-activated lime-Pozzolan/slag cement
- Alkali-activated calcium aluminate blended cement
- Alkali-activated Portland blended cement
Pacheco-Torgal et al. (2008) reported that blast furnace slag with sodium hydroxide as an
activator and process developed in two stages. The liberation of silica aluminium and cal-
cium hydroxide took place, forming of silica and alumina hydrate occurred as the regeneration
of the alkali solution. A number of researchers (Pacheco-Torgal et al., 2008; Roy, 1999; Shi
et al., 2011) reported that the difference between the compositions of traditional Portland
cement and fundamental rock-forming minerals of the earth crust. The calcium silicate hy-
drate (CSH) and calcium hydroxide (Ca(OH)2), tthe major reaction products, are containing
alkalis. The formation of major reaction products is raised by the probability of enhanced
durability. Based on investigations, it was developed as “soil-cement” which was a new type
of binder (Glukhovsky et al., 1980). TThe soil-cement was found from ground aluminosili-
cate mixed with rich alkalis industrial wastes. The general mechanism of alkaline activation
reaction of material into three stages, destruction-coagulation, coagulation-condensation, and
condensation-crystallisation, introduced by Glukhovsky (Shi et al., 2011). Recently, several
authors (Criado et al., 2005; Duxson, Fernandez-Jimenez, Provis, Lukey, Palomo and Van De-
venter, 2007; Shi and Fernandez-Jimenez, 2006) extended Glukhovky theories to explain the
polymerisation process. Figure 2.2 shows descriptive polymerisation mechanism of AAC. The
most important key process of the reaction mechanism of AAC is the transforming of alumi-
nosilicate source into a synthetic alkali aluminosilicate by fine grinding and heat treatment.
Rangan (2008) recommended heat treatment of AAC mixtures as being substantially assistant
to the chemical reaction of aluminosilicate source to form a synthetic alkali aluminosilicate.
11
Figure 2.2: Descriptive model for alkali activation of aluminosilicate (Shi et al., 2011)
Generally, the high temperature accelerates the polymerisation process compared to the am-
bient condition. Hardijito and Rangan (2005) indicated that alkali activated fly ash cement
(AAFA) did not harden immediately at room temperature. They reported that when the room
temperature was less than 30◦C, the hardening did not occur at least for 24 hours. AAFA pro-
duced in ambient temperature achieved lower strength in the early days as compared to heat
treated specimen. Kong and Sanjayan (2008) reported that longer heat curing regime did not
significantly affect the strength behaviour of AAFA. They found that most of polymerisation
were complete within the first 24 hours of heat curing.
Alkali activator solution is an important constituent in AAC. Alkaline activators produce a
dissolution of these source of materials. The alkaline condition activates the reaction between
raw materials in polymerisation process to form the polymer network. The aluminosilicate
materials can increase reactivity as a lower bonding energy of Al-O than Si-O. The alkali
activator breaks down the covalent bond Si-O-Si and Al-O-Al in aluminosilicate materials with
pH of alkali activator as follows (Shi and Day, 2000):
≡ Si−O − Si ≡ +3OH− → (SiO (OH)3)−
(2.3.1a)
≡ Si−O −Al ≡ +7OH− → (SiO (OH)3)−
+ (Al (OH)4)−
(2.3.1b)
The dissolved mono-silicate and aluminate form amorphous to semi-crystalline silico-aluminate
structure and its empirical formula is (Caijun and Della, 2006):
Mn (− (Si−O2)z −Al −O)n· wH2O (2.3.2)
where M is a cation such as sodium, n is the degree of ploy-condensation, and z is a poly
chain and ring polymers with oxygen; range from amorphous to semi-crystalline. However, the
reaction mechanism of AAC is still not fully understood.
The ratio of pozzolan to alkaline activator is critical in strength development and thermal
resistance of AAC (Kong and Sanjayan, 2008). The most common alkaline activators in AAC
are potassium hydroxide (KOH) with potassium silicate and sodium hydroxide (NaOH) with
sodium silicate (Na2SO4). It was proven that alkaline solution containing soluble silicate could
increase reactivity compared to alkali solution containing only hydroxide (Palomo et al., 1999).
Hardijito and Rangan (2005) reported the alkaline solution containing soluble silicate, either
12
sodium or potassium silicate, had a high rate of reactions than that of alkaline hydroxides. In
addition, Xu and Van Deventer (2000) studied alkaline solution by adding Na2SO4 solution to
NaOH solution and found that it increased the reaction between the source materials. They
concluded that typically the NaOH solution has the higher extent of dissolution of minerals
than the KOH solution.
Rangan (2008) studied that failure behaviour and elastic properties of AAFA (Class F) concrete
are similar to OPC concrete. Also, he found that AAFA concrete has excellent compressive
strength, resistance to sulfate attack, good acid resistance, low drying shrinkage, and creep
behaviours. Hardjito et al. (2004) studied AAFA (Class F) concrete with varying concentration
of NaOH solution in morality, Na2SO4 to NaOH ratio, curing temperatures, and curing times.
They found that higher concentration in term of the morality of NaOH solution and Na2SO4
to NaOH ratio resulted in higher compressive strength. An increase of curing temperatures in
the range of 30 to 90◦C leads an increase of compressive strength. In term of curing times, the
increase in strength beyond 48 hours is not significant.
AAC has superior properties compared to ordinary Portland cement (OPC) (Demirel and
Kelestemur, 2010; Duxson, Provis, Lukey and Van Deventer, 2007; Kong and Sanjayan, 2010;
Xu and Van Deventer, 2000). Several properties which are superior to those in OPC are such
as less drying shrinkage, higher early age strength, and resistance to high chemical and tem-
peratures. Also, one of main source of OPC is CaO, which can be presented in the form of
carbonates (CaCO3). However, AAC does not contain CaCO3 because one of the main sources
is the aluminium. Therefore, alkali-activated pozzolan cement does not release vast quantities
of CO2 which can be described as due to the release of CaO from CaCO3. The application of
AAC has a wide range in the field of industries because of its superior performance and dura-
bility properties. According to Davidovits (1994a), the classification of the type of application
of AAC is described base on Si : Al ratio. A high ratio of Si : Al more than 15 provided a poly-
meric character to the polymeric material and low ratio of Si : Al of 1,2, or 3 (Rangan, 2008).
Alkali activator solution is an important constituent in AAC. The alkaline condition activates
the reaction between raw materials in polymerisation process to form the polymer network.
The ratio of pozzolan to alkaline activator is critical in strength development and thermal re-
sistance of AAC (Kong and Sanjayan, 2008). The most common alkaline activators in AAC
are potassium hydroxide (KOH) with potassium silicate and sodium hydroxide (NaOH) with
Table 2.5: Application of Geopolymer material based on silica to alumina atomic ratio (Rangan,2008)
Si:Al ratio Application
1Brick, CeramicsFire protection
2Low CO2 cement and concreteRadioactive and toxic waste encapsulation
3
Fire protection fibre glass compositeFoundry equipmentHeat resistant composites 200◦C to 1000◦CTooling for aeronautics titanium process
>3Sealant for industry 200◦C to 600◦CTooling for aeronautics SPF aluminium
20 to 35 Fire-resistant and heat resistant fibre composite
13
sodium silicate (Na2SO4). It was proven that alkaline solution containing soluble silicate could
increase reactivity compared to alkali solution containing only hydroxide (Palomo et al., 1999).
Hardijito and Rangan (2005) reported the alkaline solution containing soluble silicate, either
sodium or potassium silicate, had a high rate of reactions than that of alkaline hydroxides. In
addition, Xu and Van Deventer (2000) studied alkaline solution by adding Na2SO4 solution to
NaOH solution and found that it increased the reaction between the source materials. They
concluded that typically the NaOH solution has the higher extent of dissolution of minerals
than the KOH solution.
2.4 Alkali-Activated Lime-Pozzolan Cement
Lime Pozzolan cement (LPC) is one of the earliest building materials, widely used in the ma-
sonry construction during Roman times. Slow strength development with ambient temperature
curing condition is one of main disadvantage of lime-pozzolan cement. In the 19th century,
its use was significantly reduced because of the invention of Portland cement which was faster
setting and had high early strength. In the past 50 years, some developed countries have been
used for manufacturing construction products because of low cost and excellent durability than
Portland cement. Also, pozzolans are used for sustainability or mixing with Portland cement
for advantageous properties which are a reduction in cost and heat evolution, alkali-aggregate
expansion control, increased chemical resistance, reduced concrete drying shrinkage, improve-
ment of the properties of fresh concrete and decreased permeability (Caijun and Della, 2006).
Alkali activator is potentially possible to accelerate the early strength development. Shi (1992)
suggested that Na2SO4 is a most effective activator of LPC and 4% of Na2SO4 is the optimum
content as shown in Figure 2.3. Also, he determined that alkali activator in LPC paste can
increase early strength development (Shi and Day, 1993b) as shown in Figure 2.4. The following
products are recognised as the main hydration products of lime-pozzolan cement (Caijun and
Della, 2006).
Ca(OH)2 + Pozzolan + Water→ Hydration Products (2.4.1)
where hydration products are:
- Calcium silicate hydrate (C-S-H)
- Ettringite (C3A·3CaSO4 ·32H2O)
- Hydrated tetracalcium aluminate (C4AHx)
- AFm (C3A·CaSO4 ·12H2O)
- Hydrated gehlenite (C2ASH8)
- Hydrated calcium carbonaluminate (C3A·CaCO3 ·12H2O)
In lime-pozzolan cement, Ca(OH)2 is mixed with pozzolan in water without activator, the
solution reaches a high pH value which is approximately 12.5 at 20◦C.
Ca(OH)2 → Ca2+ + 2OH− (2.4.2)
In a high pH solution under OH−, pozzolan is dissolved and depolymerised into the solution
such as Ca2+, K+, and Na+. As Ca2+ ions which are in contact with those depolymerised or
dissolved pozzolan mono silicate and aluminate species, calcium silicate hydrate (C-S-H) and
14
Figure 2.3: Effect of Na2SO3 dosage on strength development of LPC paste (Shi and Day, 1993b)
Figure 2.4: Relationship between the Ultimate strength and curing temperature (Shi and Day,1993a)
calcium aluminate hydrate (C4AH13) form.
Y (SiO (OH)3)−
+XCa2+ + (Z −X − Y )H2O + (2X − Y )OH− → CxSyHz (2.4.3)
2 (Al (OH)4)−
+ 4Ca2+ + 6H2O + 6OH− → C4AH13 (2.4.4)
Lime-pozzolan cement has not shown significant strength after 3 days because dissolved monosil-
icate species diffuse more quickly than dissolved aluminate spices.
2 (Al (OH)4)−
+ 3SO2−4 + 6Ca2+ + 4OH− + 26H2O → C3A · 3CaSO4 · 32H2O (2.4.5)
6Ca2+ + 2 (Al (OH)4)−
+ 3CaO ·Al2O3 · 3CaSO4 · 32H2O + 10OH− (2.4.6)
→ 3 (3CaO ·Al2O3 · CaSO4 · 12H2O) + 5H2O
The reaction between Ca(OH)2 and Na2SO4 as Na2SO4 is added as expressed by (Litvan, 1986).
Na2SO4 + Ca (OH)2 + 2H2O → CaSO4 · 2H2O ↓ +2NaOH (2.4.7)
The components of quick lime contents more than 90% of Calcium Oxide (CaO). Thus, quick
lime with water would renovate into Ca(OH)2 and hydrated lime contents greater than 90%
of Ca(OH)2. Therefore, Na2SO4 act with Ca(OH)2 than transforms into Sodium hydroxide
(NaOH). Also, hydration products with Na2SO4 paste of AFt and AFm are as shown (Caijun
15
and Della, 2006):
2 (Al (OH)4)−
+ SO2−4 + 4Ca2+ + 6H2O + 4OH− → C3A · CaSO4 · 12H2O (2.4.8)
6Ca (OH)2 + 2Al2O3 + 3CaO ·Al2O3 · 3CaSO4 · 32H2O → 3 (3CaO ·Al2O3 · 12H2O) + 2H2O
(2.4.9)
Helmuth (1983) studied complete theoretical reaction between Limes to pozzolan ratio according
to the hydration products of lime-pozzolan cement. They assumed that
- Fly ash contains SiO2 = 50%, Al2O3 = 30%, Other compounds = 20%
- Average molar ratio of CaO/SiO3 of C-S-H is 1
- CaO/SiO3 weight ratio in Gehlenite hydrate (Ca2Al(AlSiO7)) is 0.55
- Additional lime required to produce AFt and AFm
It was concluded at least 45% hydrated lime should contain in lime-pozzolan cement as shown in
Figure 2.5. However, a rise in lime content will increase the water requirement in lime-pozzolan
cement and increasing the water will reduce the strength of harden paste. Over saturated
Ca(OH)2 in lime-pozzolan cement also will produce calcium carbonation (CaCO3) in paste
(Shi, 1996). The theoretical solubility of Ca(OH)2 in water is 0.178g/100mL at 20◦C. It is
impossible practically soluble Ca(OH)2 in water for cement mixture. Therefore, according to
the results of Shi and Day (1993b), 20% hydrated lime mixed with 80% of pozzolan was an
optimum mixture for lime-pozzolan cement.
Figure 2.5: Effect of Lime content on Strength development of Lime-pozzolan cement cured at 50◦C(Shi and Day, 1993a)
2.5 Characteristic of Binding Cementitious Products
There are several of techniques for characterising the composition and structure of the re-
action or the hydration products of binding cementitious materials, such as X-ray diffraction
(XRD), Scanning Electron Microscope (SEM), thermogravimetric analysis (TGA), nitrogen ad-
sorption/desorption (NAD), mercury intrusion porosimetry (MIP), and transmission electron
microscopy (DOH). Especially, XRD and SEM are widely used to identify the characteristics
of composition and structure of reaction or hydration products of binding cementitious mate-
rial because these are a rapid analytical technique. X-ray diffraction (XRD) provides evidence
of the periodic atomic structure of crystals by reflecting X-ray beams at certain angles of in-
cidence. The cement phase composition influences the performance characteristics of them
16
(Stutzman, 1996). XRD is a direct method for qualitative and quantitative characterisation of
fine-grained materials like cement and raw materials. Products in each phase have a unique
diffraction pattern that is independent of others and the intensity of each pattern is relative to
that phase concentration in the mixture. For Quantitative X-ray diffraction analysis (QXDA),
the most common method of analysing chemical composition in cement involves the addition
of a known amount of an internal standard having a controlled particle size. This internal
standard helps correct any matrix effects between specimens and time-dependent changes in
X-ray intensity (Stutzman, 1996). Scanning Electron Microscope (SEM) releases a beam of
electrons through the specimens and measures any signals resulting from the electron beam
interaction with the specimens. The images of topography could be used to study particle size,
shape, surface roughness and hydration products (Stutzman, 2001). In addition, back-scattered
electron (BSE) imaging has illustrated its potential in a study of cementitious materials. BSE
are electrons from the incident beam which are scattered over large angles that reflect from the
specimens (Scrivener, 2004).
According to (Zhang and Scherer, 2011), these techniques must be subjected to treatment
that arrests the reaction or the hydration products of binding cementitious materials. There
are two main methods for arresting hydration and removal of water from the samples, direct
drying and solvent exchange methods. In the direct drying methods, such as oven drying,
microwave drying, vacuum drying, freezing drying, D-drying and P-drying, water is removed
by convening into vapour. Microwave drying can be used for investigating total water content
and for accelerated curing but a rapid thermal extension of the pore liquid could cause damage.
D-drying, or dry ice method, is commonly recognised as the best standard drying technique. It
is assumed to remove all of physically unbound water in the cement paste pore and is considered
the best preserving microstructure drying method. The vacuum drying is generally done in the
chamber with less than 0.1 Pa pressure. It gives a similar result to D-drying method but
vacuum drying method cannot effectively arrest early hydration because the drying process is
slow. Freezing drying method causes less damage compared to oven drying method but it still
damages the microstructure and pore structure in the cement. Oven drying is probably the
simplest technique, thus widely used as the drying technique with temperature generally set
between 60 to 105 ◦C, and an atmospheric pressure of 101kPa, which will effectively remove
evaporable water. Oven drying at 105◦C for 24 hours will remove water from hydration process.
However, the oven dry method could damage the sample caused by the remaining structural
water and pore structure.
Solvent exchange method is to exchange water with isopropyl alcohol in binding cementitious
materials and most researchers (Zhang and Scherer, 2011; Scrivener and Nonat, 2011; Chen
et al., 2014) recommended this method for arresting reaction or hydration products. For X-
ray diffraction test, the presence of organic material may not matter, but the possibility of
decomposing or altering hydrated aluminate phases must be considered. The solvent to sample
ratio was suggested to be taken as 100:1 and oven or vacuum dry for 24 hours, the minimum
soaking time (t) also could be calculated by Equation , in which C0 is initial concentration, C1 is
constant surface concentration, c (r, t) is the average concentration in sample, a is characteristic
dimension, and D is diffusivity (m2/s) (Zhang and Scherer, 2011). The soaking time can be
solved using Equation , approximately 24 hours required for a soaking time when the time to
replace the water to 99.99% with a solvent in cementitious materials, 1mm of the granule of
sample size, and 1.0 × 10−11 m2/s of diffusivity of solvent assumed. Thus, binding cementitious
materials can arrest their reaction or hydration products using solvent exchange method with
17
24 hours soaking time.
c (r, t)− C1
C0 − C1=∞∑n=1
6
n2π2exp−n2π2Dt
a2≈ 6
π2expπ2Dt
a2(2.5.1)
2.6 Engineering Statistics Analysis
Civil Engineering experiments are performed to determine a characteristic of systems such as
materials and structures. These characteristics, thus, require to be evaluated appropriately
according to the objectives of the experiments which generally are to (Park, 2007):
- Evaluate characteristic values without statistical analysis.
- Identify experimental factors that are significant to the response and determine how large
the impact is.
- Determine statistically the affecting factors with small influence.
- Determine significant factors and then evaluate optimisation condition.
Design of Experiments (DOE) is a method to satisfy experimental objectives and various factors
that can be determined by statistical analysis of experimental results. Generally, the process
of DOE is as shown in Figure 2.6 and some terminologies of DOE are presented in Table 2.6.
DOE can be used by meta-models which are the model of a model. The approximated function,
which is a meta-model, can represent specific function values as vector b as follows:
DOE can be regarded by the approximated function y with vector b as:
y = g (b) (2.6.1)
Figure 2.6: General process of design of experiment (Park, 2007)
Table 2.6: Terminologies of design of experiment
Terminologies Description
Factor Influence the characteristic function or objective functionLevel Factor or design variable
Characteristic Response of the system
18
While the error in the approximation process is ε, the relationship between y and y is:
y = y + ε (2.6.2)
In DOE, a method is employed to define y and variable vector b is obtained (Park, 2007).
Therefore, DOE can be utilised to develop efficient experiments and for analysis of experimental
results.
Genichi Taguchi developed DOE method during the 1950s (Ranjit, 1990), and is now commonly
known as Taguchi’s method. Taguchi’s DOE approach to parameters design provides the design
engineer with a systematic and efficient method for investigating optimum design parameters
according to the required performance and cost. Taguchi’s DOE method identifies the ‘signal to
noise (S/N)’ factors. S/N factors are what causes a measurable product or process characteristic
to deviate from its target value (Ozbay et al., 2009). Target value might be:
Smaller is better: select when the goal is to minimise the response. The S/N ratio can be
determined as:
S/N = −10× log10
(1
n
n∑i=1
Y 2i
)(2.6.3)
Larger is better: select when the goal is to the maximum the response. The S/N ratio can be
determined as:
S/N = −10× log10
(1
n
n∑i=1
1
Y 2i
)(2.6.4)
Nominal is better: select when the goal is to target the response and it is required to base the
S/N ratio on standard deviations only. The S/N ratio can be determined as:
S/N = −10× log10
(1
n
n∑i=1
(Yi − Y0)2
)(2.6.5)
Yi is the measured value of each response in Equations (2.6.3 to 2.6.5) . When variability
occurs, it is because the physically active in the design and environment that promotes change
(Ozbay et al., 2009). S/N factors can be classified into:
- External noise factors, sources of variability that come from outside the samples.
- Unit to unit noise, due to the fact that no two manufactured components or products are
ever exactly alike.
- Internal noise, due to deterioration, ageing and wear incurred in storage and use.
Taguchi’s DOE requires creating a set of tables of numbers. These tables, known as ‘orthogonal
array’, are used to lay out particular factors and levels of constituents. Results from the
experiements are generally analysed using Analysis of Variance (ANOVA), which is to determine
the extent to which the effect of an independent variable is on the targeted tests. It is a general
procedure for isolating the source of variability in a set of measurements.
2.7 Chapter Summary
Review of cementitious materials and their binding cementitious products presented in this
Chapter outlined the chemical compositions and their reaction products. Calcium Silicate Hy-
drate (CSH) which governs fundamental properties such as strength, is important to binding sys-
tem of OPC based materials and needs attention to be fully understood. Due to beneficial uses
19
of fly in construction materials, its effects on the mechanical and chemical and microstructural
properties of hydrated cement systems are necessary to be explored. In particular, this research
will determine the possibility of establishing interactive relationships between constituents of
mixtures. Similarly, alkali-activated cement (AAC) is one of alternative cementitious mate-
rials which has superior properties such as mechanical, chemical, and thermal compared to
OPC-based binding cementitious products. However, the relationship between properties and
reaction products are also not well understood. This forms the background of this research as
further works are required to understand the reaction products and the interactive relationships
between selected constituents of mixture of AAC.
The literature review shows that the characteristics of materials and structures in civil engineer-
ing and construction require to be evaluated appropriately according to the objectives of the
experiments. Taguchi’s design experimental approach is one of DOE methods and introduced
that parameters design provides the design engineer with a systematic and efficient method
for investigating optimum design parameters for performance and cost. Analysis of Variance
can extend the knowledge on the effect of major independent variables of the data obtained
from Taguchi’s DOE. Further, use of statistical and systematical methods can help to draw
important conclusions on the properties of cementitious materials.
20
Chapter 3
Nanoindentation
3.1 Introduction
Nanoindentation is a technique used to obtain meaningful mechanical properties of materials.
This Chapter reviews a classical indentation analysis method of contact mechanism in homoge-
neous materials. This review defines the critical importance of the experimental investigation of
the fundamental properties of cementitious materials. Nanoindentation testing can determine
local mechanical properties of cementitious binder paste, mortar and concrete at a microstruc-
ture level. Indentation testing is done essentially by touching the material of interest, whose
mechanical properties such as elastic modulus, hardness, strain-hardening, fracture toughness
are unknown, using another material whose properties are known (Fischer-Cripps, 2011). The
main focus of this Chapter is to present details of analysis for determining indentation modulus
M , hardness H and related mechanical properties of indentation materials.
3.2 Homogeneous Materials
According to the literature (Cheng and Cheng, 2004; Constantinides and Ulm, 2007; Fischer-
Cripps, 1999; Oliver and Pharr, 2004), measurement of the elastic modulus and hardness of a
material can be obtained from indentation load-displacement data during one cycle of loading
and unloading. Figure 3 1 shows a contact between a rigid indenter and a flat specimen. This
figure presents behaviour of the modulus of elasticity of a material at the contact radius a, the
indenter load P , the indenter radius R, and indentation modulus M . Hertz contact equation
(Fischer-Cripps and Mustafaev, 2000) determines the indentation modulus M as:
a3 =4
3
PR
M(3.2.1)
where the quantity of M is expressed as the combined modulus of the indenter and the
specimen as:
1
M=
1− v2
E+
1− v′2
E′(3.2.2)
In the Equation (3.2.2), E is the elastic modulus of the specimen and v is the Poisson’s ratio
of the indented material, while the superscript prime denotes the corresponding properties of
the indenter, i.e., E′ is the elastic modulus and v′ is the Poisson’s ratio of the indenter. The
maximum tensile stress in the specimen occurs at the edge of the contact circle at the surface
as:
21
Figure 3.1: Schematic of contact between a rigid indenter and a flat specimen with modulus E(adapted after (Fischer-Cripps and Mustafaev, 2000))
σmax = (1− 2v)
(P
2πa2
)(3.2.3)
This tensile stress is acting in a radial direction on the outside of indentation surface. Com-
bination of Equations (3.2.2) and (3.2.3) provide the maximum tensile stress on the outside of
the contact circle, which can be expressed in term of the indenter radius R, as (Fischer-Cripps
and Mustafaev, 2000):
σmax =
(1− 2v
2π
)(4E
3
)2/3
P 1/3R−2/3 (3.2.4)
The mean contact pressure pm has the additional virtue of having actual physical significance
as:
pm =P
πa2(3.2.5)
Substituting Equation (3.2.1) into (3.2.3), the mean contact pressure pm is written as (Fischer-
Cripps and Mustafaev, 2000):
pm =
(4M
3π
)( aR
)(3.2.6)
The mean contact pressure pm is referred to as the “indentation stress”, the quantity of a/R
is the “indentation strain”. It is similar to the elastic condition of linear stress-strain response
that is commonly determined from conventional uniaxial tension and compressive test. The
mean contact pressure pm, which is defined as the “indentation hardness (H)” can be obtained
under fully developed plastic condition, which will be further described in Section 3.9.
Figure 3.2 presents a typical data set obtained with Berkovich indenter, where the parameter
P designates the load, and h the displacement relative to the initial undeformed surface. It
is important to measure the maximum load, Pmax, the maximum displacement, hmax, and
the elastic unload stiffness (or contact stiffness). Noting that S = dP/dh, which is the slope
of the upper portion of the unloading curve during the initial stage of unloading. The final
depth hr is the permanent depth of penetration after the indenter is fully unloaded which is
another important quantity. The most widely used method for determining the contact area is
developed by Oliver and Pharr (1992, 2004). The procedure for determining the contact area
begins with fitting the load-displacement curve acquired during unloading to the power-law
relation as:
22
Punloading = β · (h− hr)m (3.2.7)
where h is the penetration depth, β and m are empirically fitting parameters, and hr is the
final displacement as shown in Figure 3.2. The contact stiffness S is then established by (Oliver
and Pharr, 1992, 2004)
S = β ·m (h− hc)m−1 |h=hmax (3.2.8)
However, Equation (3.2.7) does not always provide adequate expression of the entire unloading
curves, thus, only the upper portion of the unloading around 25% to 50% of data is generally
sufficient.
3.2.1 Types of Indenter
The main achievement of nanoindentation technology is to extract elastic modulus and hardness
of materials from the load-displacement curve. Normally, the indentation hardness is related
to the measurement of the size of a residual plastic impression in the material. It provides
a measure of the contact area for a given indenter load. Thus, the penetration depth of the
specimen during indenting load is important to extract material properties of the specimen. The
known geometry of the indenter is therefore required to determine the contact area. Generally,
nanoindentation test can be conducted using several types of indenter as shown in Figure 3.3.
Spherical Indenter
The mean contact pressure pm is determined in the fully plastic zone in term of indentation
hardness H and the contact area of the spherical indenter Ac as:
Figure 3.2: Typical indentation load-displacement curve
Figure 3.3: Different types of indenter (a) Spherical (b) Berkovich (c) Conical (d) Vickers indenter
23
pm = H =P
Ac=
2P
πR2(3.2.9)
where R is the radius of the contact circle at fully loading. The contact depth as shown in
Figure 3.1 and the contact area can be calculated using the known geometry of the indenter,
thus the contact area of a spherical indenter is (Fischer-Cripps and Mustafaev, 2000):
Ac = π(2Rhc − h2c
)2(3.2.10)
Berkovich Indenter
The Berkovich indenter is a more precise control over the indentation process. The mean contact
area is normally obtained from a measure of the contact penetration depth hc as shown in Figure
3.4. The projected area of Berkovich contact is given as (Fischer-Cripps and Mustafaev, 2000):
Ac = 3√
3h2c tan2 θ (3.2.11)
where θ is 65.27◦.
Conical and Vickers Indenter
For a conical indenter, the projected area is obtained as:
Ac = πh2c tan2 θ (3.2.12)
The projected area of Vickers indenter is determined as (Fischer-Cripps and Mustafaev,
2000):
Ac = 4h2 tan2 θ (3.2.13)
3.2.2 Indentation Analysis
The indentation hardness H and modulus M of any axis-symmetric indenter can be estimated
by:
H =P
Ac(3.2.14a)
M =
√π
2
S
Ac(3.2.14b)
Figure 3.4: Berkovich geometry of contact
24
The indentation depth hr and hc are corresponding to the applied load; hr is the depth of the
residual impression, and hc is the contact depth which can be determined as:
hc = hmax − ξPmaxS
(3.2.15)
According to Oliver and Pharr (1992), the projected contact area Ac can be expressed by the
indentation depth hc as shown in Table 3.1. Sakharova et al. (2009) reported that areas of
indenters differ due to the imperfection at the tip. They presented the area function of the
indenter based on the numerical simulation of indentation process as shown in Table 3.2.
Another term in indentation analysis is the load frame compliance Cf which is the sum of the
compliance of the load frame that needs to be calibrated together with the area function of the
indenter. Oliver and Pharr (2004) developed a simple and accurate calibration procedure for
determining the total measured compliance C, which is defined as the inverse of the measured
stiffness and can be expressed by:
C = Cf +
√π
2M
1√Ac
(3.2.16)
The second term in Equation (3.2.16) is the contact which is two act like springs in series, and
the area function Ac can be determined from the measurement of the compliance as a function
of depth hc if Cf is known. The area function is proposed as (Oliver and Pharr, 2004):
Ac =8∑
n=0
Cn (hc)2−n
(3.2.17)
where Cn are constants, determined by curve fitting procedure and n is the number of constant.
Table 3.1: Project contact area, geometry factors and intercept factor for various types of indenters
IndenterProject contact
areaSemi-angleθ(degree)
Effectivecone angleα(degree)
Interceptfactor ξ
Spherical Ac ≈ 2πRhc - - 0.75
Berkovich Ac = 3√
3h2c tan2 θ 65.27◦ 70.3◦ 0.75
Vickers Ac = 4h2c tan2 θ 68◦ 70.3◦ 0.75
Cone Ac = πh2c tan2 θ α α 0.727
Table 3.2: Different indentation area function from numerical analysis (Sakharova et al., 2009)
Types of Indenter Area function A (µm2)
Berkovich A(h) = 24.675h2 + 0.562h+ 0.03216
Vickers A(h) = 24.561 (h+ 0.008)2
+ 0.206 (h+ 0.008)
Conical A(h) = 24.5 (h+ 0.011427)2
3.3 Heterogeneous Materials
Composite materials exhibit several types of heterogeneity which can be observed from the
perspective of different lengths scale. According to the literature (Constantinides et al., 2006;
Kanit et al., 2003; Pichler and Lackner, 2008; Zaoui, 2002), continuum micromechanics theory
25
offers a framework to distinguish constituents in heterogeneity. The idea of continuum microme-
chanics is a possibility of separating a heterogeneous material into phases with “on-average”
constant material properties. Continuum indentation analysis is based on spatially homoge-
neous mechanical properties. One of the most important elements of continuum approach is
the use of Representative Volume Element (RVE) to describe heterogeneous material statisti-
cally. The transition from heterogeneous material to homogeneous material can be ensured by
the characteristic size L of RVE (condition of separation) as:
d� L� (h,D) (3.3.1)
where d is the characteristic length scale of the local heterogeneities in RVE, h is the indentation
depth, and D is the characteristic size of the microstructure. Figure 3.5 illustrates a scale
transition to detect a homogeneous region of a heterogeneous material.In addition, general
nanoindentation tests are based on self-similarity that have no length scale limit. Thus, it
is important to select an appropriate indentation depth h and characteristic size L of RVE.
A good estimation is by using 3h for Berkovich indenter, and h for corner cubic, in order
to obtain an effective volume proportion (Constantinides et al., 2006). If h is much smaller
than the characteristics size of a microstructure D, i.e., h � D, then the indentation results
will yield a single phase of the material properties. On the other hand, if h is much greater
than D4, h � D, the average composite material properties will have to be obtained from
the analysed results. It should be noted that more tests (N � 1) need to be conducted
on a grid that has a spacing l larger than the effective volume proportion of the indenter
and the characteristic size of microstructure, i.e., l√N � D . This is to ensure that the
statistical analysis in each phase will present the surface fraction of the total phases. Figure
3.6 presents the important grid indentation technique in order to identify several phases of
different properties of a heterogeneous material. The grid spacing l should be defined to avoid
overlapping of each indentation impression as seen in Figure 3.6(c).
A cementitious material such as cement paste has several phases of the mechanical properties.
Jennings (2000), Tennis and Jennings (2000) proposed a model for determining two types of
calcium silicate hydrate (CSH), which are high density (HD CSH) and low density (LD CSH),
at different points of the specimen’s geometry. The content of CSH is determined in term of
its volume fraction of the indentation grid. Constantinides and Ulm (2004) determined two
types of CSH with Portlandite (CH), and clinker with nanoindentation. The result shows
that decalcification of the CSH phases is the primary source of nanometer-scale elastic mod-
ulus degradation. Nemecek et al. (2011) studied the reaction products of alkali activated fly
Figure 3.5: Element volume of material in a microscopic structure: (a) heterogeneous (b) localheterogeneous (c) homogeneous (adapted after (Dormieux et al., 2006))
26
Figure 3.6: Schematic representation of grid indentation technique: (a) average composite materialwith large indentation depth (h� D), (b) low indentation depth with one phase (h� D),(c) grid
indentation technique giving several phases using low indentation depth (h� D) adapted after(Constantinides et al., 2006))
ash-based cementitious composites (AAFA) with nanoindentation and environmental scanning
electron microscope (ESEM). Their AAFA samples were made using the ratio of activator to
solid of 0.531 and cured at 80◦C for 12 hours. They observed four peaks of the reaction products
of AAFA as:
- N-A-S-H
- Partly-activated slag
- Non-activated slag
- Non-activated compact glass (Raw fly ash particles)
The N-A-S-H phase is pure and is relative to the mechanical strength of AAFA matrix. The
contents of Si ions in N-A-S-H can be increased by the presence of Si ions in the raw materials.
It has been found that the increasing condensation degree of Si ions in N-A-S-H relates directly
to the mechanical strength gain (Fernandez-Jimenez and Palomo, 2003, 2005). The partly-
activated slag phase is intermixed with slag-like particles. The non-activated slag phase is
porous and contains non-activated slag-like particles. The non-activated compact glass phase
is solid, non-activated glass sphere.
3.4 Microporomechanics
The modulus and hardness can be representative of the particle properties such as particle
stiffness (λ), Poisson’s ratio (vs), cohesion (κ), friction coefficient (ζ) and packing density (η)
(Constantinides et al., 2006; Constantinides and Ulm, 2007; Dukino and Swain, 1992; Jennings
et al., 2007; Ulm et al., 2007):
M = λ×ΠM (vs, η, η0) (3.4.1)
H = κ×ΠH (ζ, η, η0) (3.4.2)
where ΠM and ΠH dimensionless scaling relating to stiffness and hardness. This scaling rela-
tionship can be determined by linear and non-linear microporomechanics theory. It is required
to form a hypothesis to apply microporomechanics to indentation analysis. In the present case,
the tested material is assumed to be porous. Pipilikaki and Beazi-Katsioti (2009) reported
that the porosity size of binding cementitious materials is smaller than 2.5nm. Thus, it can
27
be assumed that the characteristic size of the porosity is much smaller than the maximum
indentation depth (h) for scaling separability reason (Constantinides and Ulm, 2007).
The self-consistent or polycrystal micromechanical, model is similar to micro-elasticity of a
porous material whose solid is a granular material (Constantinides and Ulm, 2007). The self-
consistent scheme assumes that each particle of a given phase (pore or solid) reacts as if it is
embedded in an equivalent homogeneous medium (Dormieux et al., 2006). The solid percolation
threshold (η0) is the limit that the solid fraction requires for providing a continuous force path
through the system. Clear matrix-porosity morphology has continuous packing density in the
solid phase, i.e. 0 < η ≤ 1. However, a perfectly disordered porous material has a solid
percolation thread of η = 0.5 and below, i.e., if η < 0.5, then the values of the composite bulk
and shear modulus become negative, which will be considered unstable. The modulus function
is provided by the following relationship (Dormieux et al., 2006; Ulm et al., 2007):
ΠM (rs, η, η0 = 0.5) = G(9ηrs + 4G+ 3rs) (3rs + 1)
4 (4G+ 3rs) (3rs + 1)(3.4.3)
where rs = 2 (1 + vs) /3 (1− 2vs) > 0 and G is the composite bulk and shear modulus, defined
as:
G =1
2− 5
4(1− η)− 3
16(2 + η) +
1
16×[144 (1− rs)− 480η + 400η2 + 408rsη
2 + 9rs (2 + η)2]1/2 (3.4.4)
For solid Poisson’s ratio of vs = 0.5, Equation (3.4.4) becomes a linear scaling of modulus with
packing density (η):
ΠM = 2η − 1 ≥ 0 (3.4.5)
The hardness with packing density (η) for the scaling relation is introduced by non-linear
micromechanics:
ΠH (ζ) = Π0 ×[1 + (1− η) ζ − (d− eη) ζ2 − (f − gη) ζ5
](3.4.6)
where Π0 is a function of frictionless portions as:
Π0 =12η (a− bη) [(2η − 1) (2 + η)]
1/2
(1− cη) (2 + η)(3.4.7)
Based on microporomechanics, the coefficients suitable for Berkovich indenter are: a = 0.19567,
b = 0.03739, c = 0.77999, d = 20.3138, e = 31.5352, f = 52.1817 and g = 99.3465 with η0 = 0.5
(Ulm et al., 2007).
An inverse application to determine the solid properties and packing density is introduced by
Ulm et al. (2007) and Bobko and Ulm (2008). In an inverse application, the solid properties (λ,
vs, κ, ζ) and local packing density (η) are unknowns. An algorithm implemented as a quadratic
minimization problem between the experimental values and the theoretical scaling relationship
can be written as:
min
N∑i=1
[Mi − λ×ΠM (vs, ηi)− (Hi − κ×ΠH (ζ, ηi))]1/2
(3.4.8)
where N is the total number of the indentation tested points, Mi and Hi are as obtained from
Equations (3.4.1) and (3.4.2). Solving Equation (3.4.8) will provide the analytical results of the
28
scaling relationship between the indentation modulus and hardness versus the packing density
of the tested material.
Mercury intrusion porosimetry (MIP) is a common measurement for providing a valid estimation
of the pore size distribution of porous solid (Diamond, 2000). The measurement of porosity
with MIP, however, appears to be valid with limited application and difficulty in estimating
the porosity of cementitious based material. A way to determine porosity with statistical
indentation technique is by the new nonintrusive way (Ulm et al., 2007). Therefore, it is
possible to calculate the total porosity of the cementitious material from:
φ =N∑j=1
fj (1− nj) (3.4.9)
3.5 Deconvolution Technique
Heterogeneous materials or composite materials are generally complex. Thus, a statistical
technique is used for analysis of indentation results. It is computationally more convenient to
deconvolute the cumulative distribution function (CDF) rather than the probability density
function (PDF) (Ulm et al., 2007). The reason for using CDF rather than PDF is that the
generation of the experimental PDF usually requires a choice of bin size for histogram construc-
tion. Thus, the deconvolution technique starts with the generation of the experimental CDF.
In the following equations, let X be a type of the mechanical properties and, as before, N is
the number of indentation tests conducted on the specimen. The experimental CDF can be
written as:
Φe(X) =i
N− 1
2Nfor i ∈ [1, N ] (3.5.1)
in which, the following Gaussian distribution gives the theoretical CDF for each phase:
Φtj(X) =1
σj√
2π
∫ Xj
∞exp
(− (s− µi)2
2σ2j
)dt for i ∈ [1, N ] (3.5.2)
where µj is the mean, σj is the variance of the distribution and j is a phase present in the
sample. The theoretical CDF function can be found by minimising the following error function:
min
N∑j=1
∑(fiΦ
tj (X)− Φe (X)
)2(3.5.3)
Moreover, the total volume fraction can be expressed by a summation of all the single phases
fj as:
n∑j=1
fj = 1 (3.5.4)
where n is the total number of phases identifiable from the analysis. To avoid the overlapping
of each phase, the constraint of the minimisation problems is required as:
µi (X) + σj (X) ≤ µj+1 (X) + σj+1 (X) (3.5.5)
The mean and standard deviation of the results from the deconvolution technique can then be
used to determine the mechanical properties and surface volume fraction of each phase.
29
3.6 Indentation Fracture Toughness
Nanoindentation data can be used to determine the fracture toughness of materials to measure
the radial and lateral cracks indentation points (Chen and Bull, 2007; Ling, 2011). A measure
of fracture toughness using radial and lateral cracks is based on fracture mechanics. According
to Lawn et al. (1980), the indentation fracture toughness is formulated as:
Kc = ε
(E
H
)nP
c3/2(3.6.1)
where ε is empirical calibration constant, c is the radial crack length. A number of researchers
(Anstis et al., 1981; Dukino and Swain, 1992; Field et al., 2003; Laugier, 1987) conducted to
determine ε and n terms in Equation (3.6.1) using indentation test. It can be estimated from
indentation with Berkovich indenter that:
Kc = 1.073xv
( aL
)1/2(E
H
)2/3P
c3/2(3.6.2)
where Kc is fracture toughness, xv is 0.015, and a and L are measured by radial cracking length
as shown in Figure 3.7.
The indentation P−h curve can also be represented by a consideration of the energy transferred
during loading and unloading (Oliver and Pharr, 1992). The area under the loading (Ploading)
and unloading (Punloading) versus the indentation depth (h) curves represents the plastic energy
(Wp) and elastic energy (We), as shown in Figure 3.8. The total energy (Wt) is the sum of Wp
and We, or the total area under the P − h curve. Oliver and Pharr (1992) suggested using the
loading and unloading curve fitting with a power law function.:
Ploading = α · h2 (3.6.3a)
Punloading = β · (h− hr)m (3.6.3b)
where α, β and m are constants. The variation of m in the range 1.2 ≤ m ≤ 1.6, i.e., m = 1
for flat punch, m = 1.5 for parabolic and m = 2 for cone intender tip. By fitting loading and
unloading curve with a power law function, Equation(3.6.3), Wt and We can be obtained from:
Wt =
∫ hmax
0
Ploadingdh (3.6.4)
We =
∫ hmax
hr
Punloadingdh (3.6.5)
Figure 3.7: Berkovich indentation crack parameters (adopted after (Ling, 2011))
30
Figure 3.8: Indentation load (P ) and depth (h) curve represent energies (adapted after (Cheng andCheng, 2004))
The relationship between Wp/Wt and hr/hmax can be expressed as (Cheng and Cheng, 2004;
Chen and Bull, 2007):
Wp
Wt= 1−
[1− 3 (hr/hmax)
2+ 2 (hr/hmax)
3
1− (hr/hmax)2
](3.6.6)
The total indentation work Wt can be determined from:
Wt = Wp + Ufracture +We +Wother (3.6.7)
where Ufracture is the fracture dissipated energy, and Wother is another energy transferred such
as heat energy dissipation or thermal drift. However, in the present research, Wother from
the thermal drift is ignored in the calculation of Wt. The critical energy (Gc) and fracture
toughness (Kc) can be expressed as (Taha et al., 2010):
Gc =∂Wfracture
∂A=Ufracture
Ac(3.6.8)
Kc = (GcM)1/2 (3.6.9)
The approach will be used in determining fracture toughness based on the assumption that the
crack growth under indentation loading is stable.
3.7 Finite Element Analysis
The finite element (FE) method is commonly used for modelling different engineering problems.
According to some literature (Bhattacharya and Nix, 1991; Lichinchi et al., 1998; Stauss et al.,
2003), FE simulation could be an appropriate method for determining indentation properties.
Knowing that purely elastic deformation takes place only during the beginning of the indenta-
tion process, the Von-Mises yield criterion is applied to determine the occurrence of the plastic
deformation. The equivalent Von-Mises stress is given by:
31
σVM
=
√(σ1 − σ2)
2+ (σ2 − σ3)
2+ (σ3 − σ1)
2
2(3.7.1)
where σ1, σ2 and σ3 are the three principal stresses. When σVM reaches the yield strength σy,
the material begins to deform plastically. For the conical indentation in an elastic-perfectly
plastic behaviour material, indentation pressure is written as(Gaillard et al., 2003):
pm =2
3σy
[1 + ln
(M · tanα
3σy
)](3.7.2)
where pm is the indentation pressure (i.e. pm = Pmax/Ac), M is indentation modulus of the
sample and α is the conical indentation face angle, as already presented in Table 3.1. The
elastic-perfectly plastic behaviour is governed by the elastic and yielding properties of the
material.
3.8 Time-depending Nanoindentation
It is well known that materials exhibit viscous and elastic behaviours under deformation. Vis-
coelastic materials store energies under deformation, return to its original state upon removal
of stress, and can be under the influence of time-dependent stress-strain factors. Nanoindenta-
tion can be used to determine quantitative viscoelastic properties. Previously, the method of
analysis of nanoindentation was assumed that behaviours of material are in an elastic-plastic
manner. However, time-dependent viscoelastic properties, such as creep and relaxation, can
occur under indentation impressed (Fischer-Cripps and Mustafaev, 2000; Ling, 2011), therefore
it is necessary to assess these properties. The relationship between indentation load (P ) and
penetration depth (h) is derived as (Sneddon, 1965; Vandamme et al., 2012):
P (t) = γMoh(t)1+1/n (3.8.1)
where γ is:
γ =2
(√πB)
1/n
n
n+ 1
[Γ (n/2 + 1/2)
Γ (n/2 + 1)
]1/n
(3.8.2)
B can be defined by z = Brn, as shown in Figure 3.9 and Γ is the Euler Gamma function given
as:
Γ (X) =
∫ ∞0
tx−1 exp(−t)dt (3.8.3)
The instantaneous bulk (Ko) and instantaneous shear modulus (Go) are related to the instan-
Figure 3.9: Indentation geometry probe
32
taneous material modulus (Eo) and instantaneous modulus (Mo), given as:
Mo =Eo
(1− v2)(3.8.4a)
Go =Eo
2 (1 + v)(3.8.4b)
Ko =Eo
3 (1 + 2v)(3.8.4c)
where v is the Poisson’s ratio of the tested material. The coefficients n and B of the indentation
geometry in Equation (3.8.2) are defined according to the indentation shape as listed in Table
3.3.
As viscoelastic behaviour is the role played by time, if the material is ideally viscous fluid, the
stress can be instantaneously infinite under constant strain. However, actual material behaviour
shows that under constant strain, the stress generally decreases from its initial value rapidly and
later more gradually as shown in Figure 3.10 (Pipkin, 2012). This behaviour is known as stress
relaxation. Another important behaviour is creep, i.e., deformation of viscoelastic materials
under constant stress increases with time. Consider a rigid indenter of axisymmetric shape
impressing an infinite half-space is made of no aging linear viscoelastic material. The functional
Equation (3.8.1) can be applied to obtain the solution to a linear viscoelastic indentation in
the Laplace domain from the linear elastic solution by replacing the elastic constants in the
elastic solution. Figure 3.11 illustrates conventional relaxation and creep stress with time, which
present the behaviour of general viscoelastic material.
According to Riesz theorem (Stein, 1956), if the functional J is linear and equicontinuous which
is the creep function, strain (ε) is given as:
ε(t) =
∫ t
τ0
J(t− τ)dσ(τ) (3.8.5)
where τ0 is the time at rest without stress and strain, and τ is the time domain. Equation
(3.8.5) can be re-written by inverse relations as:
Table 3.3: Indentation geometry
Indenter n B
Cone or Berkovich 1 cot θ
Sphere 2 1/ (2R)
Flat punch →∞ 1/ (an)
Figure 3.10: Viscoelastic behaviour (a) ideal, (b) actual viscous fluid
33
Figure 3.11: Relaxation and creep stress with time
σ(t) =
∫ t
τ0
M(t− τ)dε(τ) (3.8.6)
where exchanging roles of stress and strain that gives M (t− τ) is the specific relaxation func-
tion, i.e., the stress response to a unit step of strain (Marques and Creus, 2012). In this case,
Equation (3.8.1) can be applied using Riesz theorem with indentation load and displacement
relation in time domain as:
P (t) = γ
∫ t
τ0
M (t− τ)dh
1/1 + n (τ)
dτdτ (3.8.7)
For relaxation test, the displacement can be described via a Heaviside step function:
h(t) = hmaxH (t) ; H (t) =
{0 for t < 0
1 for t ≥ 0
}(3.8.8)
Substituting Equation (3.8.8) into Equation (3.8.7) yields:
P (t) = γ
∫ t
τ
M (t− τ)h1/1 + n
max δ (τ) dτ (3.8.9)
where δ is Dirac delta function. This equation can be obtained in the Laplace domain as:
L{P} (s) = γsL{M} (s)h1/1 + n
max (3.8.10)
Moreover, application of the standard Laplace table yields in P (t) in the time domain:
P (t) = γM (t)h1/1 + n
max (3.8.11)
Thus, the contact relaxation modulus at any time t, Mc (t), can be determined as:
Mc (t) =1
γh1/1 + n
max
P (t) (3.8.12)
In this case, the functional Equation (3.8.1) also leads to the following time-dependent inden-
34
tation displacement under Riesz theorem as:
h (t)1/1 + n
=1
γ
∫ t
τ0
J (t− τ)dP (t)
dτd (τ) (3.8.13)
Heaviside step loading P (t) = PmaxH (t) is then applied to Equation (3.8.13), resulting in:
h (t)1/1 + n
=1
γ
∫ t
τ0
J (t− τ)Pmaxδ (τ) d (τ) (3.8.14)
This equation can be obtained in a Laplace transform as:
L{h}(s)1/1 + n =1
γsL{J} (s)Pmax (3.8.15)
Applying inverse Laplace transform to this equation gives the formulation of the contact creep
compliance Jc (t) as:
Jc (t) =γ
Pmaxh (t)
1/1 + n(3.8.16)
Moreover, the indentation creep compliance is linked to the contact relation modulus in the
Laplace domain as:
[sL{M} (s)]−1
= sL{J} (s) (3.8.17)
This material behaviour is time dependent and is generally obtained in term of a linear viscoelas-
tic model (Rheological model), such as Maxwell model, three-element Kelvin-Voigt model, and
combined four-element Maxwell-Kelvin-Voigt model, as shown in Figure 3.12 (Fischer-Cripps,
2004). Ideally, in a perfect linear elastic and massless condition, helicoidal spring represents
Hooke’s Law as:
σ (t) = Eε (t) (3.8.18)
The dashpot is an ideal viscous element that can be expressed by a rate proportion of the
applied stress:
σ (t) = ηε (t) (3.8.19)
where ε = dε/dt is the rate of strain and η is the unit of viscosity.
The case of a conical indenter for Maxwell model, the time-dependent of penetration depth
with constant indentation load Po can be expressed by:
h2 (t) = γPo
[(1
Mo+
1
ηMt
)](3.8.20)
Using a similar approach, the penetration depth with time for Kelvin-Voigt model is represented
as:
h2 (t) = γPo
[(1
Mo+
1
Mv
(1− exp(−Mv
ηvt)
))](3.8.21)
In the case of combined Maxwell-Kelvin-Voigt model, the penetration depth increases with time
according to:
h2 (t) = γPo
[(1
Mo+
1
Mv
(1− exp
(Mv
ηvt
))+
1
ηM
)](3.8.22)
35
Figure 3.12: Mechanical models of time depending properties of material (a) Maxwell model (b)Kelvin-Voigt Model (c) Combined Maxwell-Kelvin-Voigt model
Equations (3.8.20) to (3.8.22) can be applied to loading whereby the contact radius increases
with time and the term in the square brackets represents the time response of the materials
(Fischer-Cripps, 2011).
3.8.1 Indentation Contact Relaxation Modulus
The relaxation modulus is obtained due to decrease in stress under constant strain as shown in
Figure 3.11. From Equation (3.8.12), the contact relaxation modulus Mc (t) can be determined
for conical indenter as (Vandamme et al., 2012; Vandamme and Ulm, 2006):
Mc (t) =π cot θ
2h2maxP (t) (3.8.23)
For spherical indenter, the contact relaxation modulus Mc (t) is:
Mc (t) =3P (t)(
4√Rhmax
)3/2P (t) (3.8.24)
This equation can be used to obtain instantaneous modulus Mo, which is (Mc (t = 0)). However,
practically, the measurement of the indentation contact relaxation modulus in Equation (3.8.23)
or Equation (3.8.24) may be underestimated because of the effect of a sharp indenter tip such
as Berkovich on the time-independent and instantaneous plastic deformation. As a result, the
contact relaxation modulus relations take into account the occurrence of plasticity during the
loading phase, the measured load relaxation response P (t) can then be linked to the indentation
contact relaxation modulus function (Cao et al., 2010; Vandamme et al., 2012). Therefore,
normalised contact relaxation modulus (Mc (t)) is defined as the relaxation modulus divided
by instantaneous modulus (Mo) and depends only on the relaxation load, thus, the contact
relaxation modulus rate is:
Mc (t) =Mc (t)
Mo=P (t)
Pmax(3.8.25)
Prony series can describe the contact relaxation modulus as (Cao et al., 2010):
36
Mc (t) = M∞ +N∑i=1
Ri exp
(−tτMi
)(3.8.26)
where Ri is the relaxation coefficient and τi is relaxation time.
The relaxation modulus Mc (t) is greatly significant in engineering applications. Construction
materials with energy dissipation ability are required for vibration reduction and hazard miti-
gation of civil engineering structures. It is important to determine the damping behaviour of
construction materials, as it is generally required to determine the loss factor which is defined
by the ratio of the loss modulus Mcl to the storage modulus Mcs at a given frequency ω (Cao
et al., 2010; Nashif et al., 1985). The loss factor, tan δ, can be determined as:
tan δ =Mcl
Mcs=
N∑i=1
giωτgi
1 + ω2 (τgi )2
(1−
N∑i=1
gi +N∑i=1
giω2 (τgi )
2
1 + ω2 (τgi )2
)−1(3.8.27)
Using Equation (3.8.27), the damping capacity, which is the energy dissipated per cycle of
motion, can be determined as:
4U = 2πUmax tan δ (3.8.28)
where Umax is maximum potential energy or maximum kinetic energy.
The contact relaxation modulus can be analysed in term of Rheological models in Figure 3.12.
In this research, a relationship is established between the contact relaxation modulus and creep
compliance, using Equation (3.8.17), and in respective of the three models. For Maxwell model,
it is proposed that the analytical contact relaxation is given as:
Manc (t) = Mo · exp
(−Mo
ηMt
)(3.8.29)
Similarly, the analytical contact relaxation modulus of Kelvin-Voigt model is proposed as:
Manc (t) =
MoMv
Mo +Mv+M2o · exp
(− (Mo+Mv)
ηvt)
Mo +Mv(3.8.30)
Also, the analytical contact relaxation modulus of Maxwell-Kelvin-Voigt model is proposed as:
Manc (t) =
MoMvηMMoηM +MvηM +MoMv
+M2o η
2M · exp
(−MoηM+MvηM+MoMv
ηv(Mo+ηM ) t)
(Mo + ηM ) (MoηM +MvηM +MoMv)(3.8.31)
3.8.2 Indentation Contact Creep Compliance
The creep properties of cementitious materials are still an enigma because the difficulty linked
to the timescale involved. Generally, the complex creep behaviour of cementitious material
is largely related to the viscoelastic response of the primary hydration or reaction products
and the binding phase of hardening. Creep testing involves applying a constant instantaneous
stress σ0 to the specimen and measuring strain as a function of time during stressing. The
creep compliance depends on the geometry of the axisymmetric indenter and maximum value
of control variables in the indentation process.
According to Vandamme and Ulm (2013, 2009), the measured indentation displacement re-
sponse h (t) is related to the contact creep compliance rate Jc (t) as:
37
Jc (t) =2a (t)
Pmaxh (t) (3.8.32)
where a (t) is the radius of the projected contact area, and h (t) is the change in the inden-
tation depth over the holding time, which can be expressed by a single power and logarithm
function as (Vandamme and Ulm, 2013, 2009):
h (t) = x1 ln
(t
x2+ 1
)+ x3t+ x4 (3.8.33)
where x1, . . ., x4 are constants. In the Equation (3.8.33), the material related term is only the
logarithmic term. Thus, h (t) can be replaced by the term of x1/t. From Equation (3.8.32), the
long-term contact creep compliance rate for short indentation creep experiment is given by:
Jc (t) =1
Ct, where C =
Pmax2acx1
(3.8.34)
where C is the contact creep modulus, and ac is the contact radius of maximum projected
area between the indenter and the indented sample just before the unloading phase, i.e., ac =√Ac/π. The higher contact creep modulus will lead to the lower logarithm creep of the material.
In Equation (3.8.32), it is required to input the radius of the projected contact area which cannot
determine directly from indentation test results. Therefore, in this research, an alternative
method is proposed by adapting to determine the contact creep compliance rate using directly
indentation load-displacement curve. Based on Equation (3.8.16), the contact creep compliance
rate for conical indenter can be described by:
Jc (t) =d
dt
(2 tan (a)
πPmaxh (t)
2
)(3.8.35)
The long-term creep compliance rate, h (t) can be replaced by x1/t, and substitute h (t) into
hmax then re-arrange equation to obtain the contact creep modulus (C) as:
Jc (t) =1
Ct, where C =
πPmax4 tan (a)hmaxx1
(3.8.36)
Similarly, the contact creep compliance rate for spherical indenter is given by:
Jc (t) =d
dt
(4√R
3Pmaxh (t)
2/3
)(3.8.37)
Thus, the long-term creep compliance is below:
Jc (t) =1
Ct, where C =
Pmax
2√Rhmaxx1
(3.8.38)
The contact creep compliance can also be described on Rheological models. The contact creep
compliance for Maxwell model is given as:
Janc (t) =1
Mo+
1
ηMt (3.8.39)
For Kelvin-Voigt model, the contact creep compliance is:
Janc (t) =1
Mo+
1
Mv
[1− exp
(−Mv
ηvt
)](3.8.40)
38
Similarly, the contact creep compliance for Maxwell-Kelvin-Voigt is:
Janc (t) =1
Mo+
1
Mv
[1− exp
(−Mv
ηvt
)]+
1
ηMt (3.8.41)
3.8.3 Numerical Analysis of Viscoelasticity
To confirm the validity of indentation viscoelastic properties in cementitious material, the com-
putation study was performed with a commercial finite element (FE) software package ANSYS
(ANSYS R○, 2015) in the present study. A 2-D axisymmetric model was employed for simula-
tion of the material behaviour in nanoindentation process as shown in Figure 3.13. A conical
indenter tip with half angle of 70.3◦, which is the effective angle of Berkovich as shown in Table
3.1. A fine mesh was defined under the contact area and near the indenter tip for a more
accurately as shown in Figure 3.14. The indented material model meshed with 4731 four-node
quadrilaterals. The contact between the rigid conical indenter and the indented sample was
considered frictionless contact. Indenter modulus was assumed 1141 GPa and Poisson’s ratio of
the indenter was 0.07. The FE indentation process was simulated with a non-linear geometry
method of displacement control.
Contact relaxation modulus
For contact relaxation modulus, the indented material was modelled using the Maxwell model
in the Equation (3.8.29) with Poisson’s ratio as 0.25 and Modulus of elasticity as 10 MPa. The
contact relaxation modulus of FE simulation and analytical and Maxwell model results are
shown in Figure 3.15. The results show that there is no difference between the numerical and
the proposed analytical results, but there is an 8.29% difference between the numerical results
and the original Maxwell model of the normalised contact relaxation modulus. This difference is
common in practice and a multiple correction factor is recommended to the Rheological model
for elastic indentation (Vandamme et al., 2012). Thus, Equation (3.8.25) is somewhat inaccu-
rate in the sense that it also fails to capture this radial contraction. As a result, indentation
relaxation modulus can be determined from nanoindentation test of cementitious material and
the time-dependent relaxation modulus can also be obtained.
Figure 3.13: A conical indenter impress on specimen
39
Figure 3.14: Meshing configuration
Figure 3.15: Normalised contact relaxation modulus with analytical, numerical solution andMaxwell model
40
Contact creep compliance
FE simulation of contact creep compliance was carried out using Maxwell creep model with
varying Poisson’s ratios as 0.25, 0.35 and 0.49. Figure 3.16 shows the results of the numerical
contact creep compliance and of using the Maxwell model. The errors between numerical and
Maxwell solutions are less than 1%. Therefore, the contact creep compliance can be determined
with indentation results.
For a logarithmic creep function, Equation (3.8.33) was used to obtain the change in indentation
displacement. With logarithmic creep function, a long-term contact creep compliance was
determined using the Vandamme Equation (3.8.34) and Equation (3.8.36), which is formulated
in the this research. Figure 3.17 shows the contact creep compliance rate based on the two
equations. It can be seen that there is a minor difference between the proposed and Vandamme
equations. For a sharp indenter such as Berkovich and conical indenter, it is hard to obtain a
projected area during the indentation process. Thus, the error between the two equations was
expected because the contact radius of the maximum projected area ac had to be assumed in
Vandamme solution, i.e., ac ≈√Ac/π because exact projected area ac physically impossible to
determine. As an alternative approach, Equation (3.8.36) is convenient to obtain the contact
creep compliance rate without knowing the contact radius of projected area ac. Therefore, the
proposed equation can determine the long-term behaviour without knowing the projected area
ac, and a nanoindentation creep experiment can determine the contact creep compliance. In
addition, the long-term contact creep compliance also can be obtained from a short period of
indentation test results.
Figure 3.16: Numerical and Rheological (Maxwell) solution of contact creep compliance
41
Figure 3.17: Contact creep compliance rate with numerical analysis
3.9 Indentation Stress-Strain Curve
Section 3.2 introduced the “Indentation stress pm” and “Indentation strain εi”. These indenta-
tion stress-strain curves can be measured by contact pressure pm, contact radius a, and indenter
radius R, as shown in Figure 3.18. The indentation stress-strain curve can be expressed by three
regimes of deformation that is the elastic, elastic-plastic and fully plastic regime. The elastic
regime is covered before the initial yield point. The region between the initial and stationary
yield point is described as the elastic-plastic regime, and over the stationary yield point is fully
plastic regime (Cao and Zhang, 2008). Determination of indentation stress pm requires indenta-
tion loading P and the projected contact area Ac with respect to the indenter tip and indented
material. It is noted that indentation stress pm also signifies the instant hardness H = P/A.
However, indentation tests cannot measure indentation contact depth hc, only measure the
maximum depth h. Therefore, it is required to determine the relationship between h and hc
as a function of hc = f (h) which depends on the indenter tip geometry and indented material.
The function hc = f (h) can be formulated by continuous measures of contact stiffness (CSM)
between indented material and indenter process. Knowing hc = f (h) and the instant contact
area of the indenter, the instant stress pm during the loading process is (Martinez et al., 2003):
pm =P
Ac(3.9.1)
The contact area function Ac can be obtained simply by curve fitting with hc as:
Ac =8∑
n=0
Cn (hc)2−n
(3.9.2)
The contact area function, in the Equation (3.9.2), is calibrated by a standard material such as
fused silica that has 72 GPa of elastic modulus and Poisson’s ratio of 0.18. The contact radius
and indenter tip radius are required to obtain indentation strain. The equivalent contact radius
42
Figure 3.18: Schematic uniaxial and indentation stress-strain curve (adopted after (Martinez et al.,2003))
a can be calculated directly from the contact area Ac as:
a =√
2Rhc − h2c for Spherical tip
a = 2√Ac/3
√3 for Berkovich tip
Cao and Zhang (2008) reported that if it is assumed that the blunt tip profile is approximately
circular, the maximum contact depth δ0, which depends on the indenter shape can be estimated
by:
δo = R (1− sin θ) (3.9.4)
where θ is semi-angle of the indenter tip. From contact mechanics analysis, the equivalent
indentation strain at shallow contact depth can be expressed as:
εi = Ka
Rwhen hc < δo (3.9.5a)
εi ≈ log
(hcδo
)when hc ≥ δo (3.9.5b)
where K is a constant, which can be determined by calibrating with a standard material. Thus,
the indentation stress-strain curve can be observed from indentation test results, and it will
display material elasticity and plasticity properties. According to Herbert et al. (2001), the
uniaxial tension stress-strain σt − εt are equivalent to indentation stress-strain pm − εI results,
and the yield strength σy can be approximated by:
σy ≈pm1.07
(3.9.6)
43
Indentation performed on the standard fused silica (E = 72 GPa) using the Spherical and
Berkovich indenter tip with CSM method can establish the validity of indentation stress-strain.
Firstly, the test was conducted using Spherical tip (150 µm of the radius). The relationship of
hc = f (h) is determined by the contact displacement hc and the instant maximum displacement
h using Equation (3.2.15). With Spherical tip, the indentation stress-strain curve on fused silica
is as presented in Figure 3.19. The result of the stress-strain curve shows that the modulus of
elasticity (initial slope) is 71.8±4.63 GPa, which is in good agreement with calibration constant
K = 7.445 ± 1.526. However the initial yield point σI
cannot be indicated, which means only
elastic regime can be observed with Spherical tip.
Similarly, the projected area function was indented with Berkovich tip (20nm of the radius)
on fused silica. The indentation stress-strain curve of fused silica is presented in Figure 3.20,
which shows elastic, elastic-plastic, and fully plastic behaviour of the material. The initial
yield point σI
is observed to be around 3.2 GPa, and the stationary yield point σII
is obtained
approximately 9 GPa. The stationary yield point σII
corresponding to its measured hardness
values was reported to be 8.5 GPa by Oliver and Pharr (1992). However, some literature (Cao
and Zhang, 2008; Martinez et al., 2003) reported that the initial yield point σI
is not clear to
observe because the fully plastic flow develops early in the loading stage with more than 20nm
of the radius of Berkovich tip.
Therefore, indentation method can provide indentation stress-strain curve, and this result can
be used to determine the yield strength of the indented material. The Spherical indenter with
150 µm of radius tip provides elastic regime only, but Berkovich tip can provide elastic to plastic
regime clearly. Martinez et al. (2003) studied the indentation stress-strain curve with different
indenters geometry to get a complete mechanical behaviour of the thin film as shown in Figure
3.21.
Figure 3.19: Indentation stress-strain curve on Fused silica with Spherical tip
44
Figure 3.20: Indentation stress-strain curve on Fused silica with Berkovich tip
Figure 3.21: The deformation regimes with different types of indenter tip (adapted after (Martinezet al., 2003))
45
3.10 Chapter Summary
The review of present tools of indentation analysis in this Chapter shows that indentation prop-
erties such as indentation modulus M and hardness H that are related respectively to elastic
and strength properties with the projected area during the loading and unloading process. This
projected area is generally estimated using Oliver-Pharr method. Based on the principle of mi-
croporomechanics, the indentation modulus and hardness represent particle properties such as
particle stiffness λ, Poisson’s ratio vs, cohesion κ, friction coefficient ζ and packing density η.
The statistical analysis tool of indentation of a multiphase material is characterised using inden-
tation of a material from the micro-scale to macro-scale. These tools for indentation analysis
of cementitious material identify the link between the multiphase compositions, and elastic,
microstructure and viscoelastic properties. The indentation process can determine indentation
stress-strain relationship of the indented material, which depends on the geometry of indenter
tip during the indentation procedure. The indentation stress-strain curves are useful for de-
termining the mechanical properties of the indented material. The analysis of nanoindentation
data presented in the Chapter has industrial and scientific benefits such as:
- Assessment of the viscous properties of the reaction products in binding cementitious
materials.
- Indentation results are comparable to conventional experimental results, such as Modulus,
hardness, relaxation modulus, creep and fracture toughens. Thus, one set of indentation
test results is representative of, and can replace, these conventional experiments.
- Indentation test identifies determinants of properties in nanoscale which enables an as-
sessment of the long-term macroscopic behaviour that is in the orders of magnitude faster
than what can be found by macroscopic analysis.
46
Chapter 4
Properties of Cement Materials
4.1 Introduction
This Chapter contains two main parts. The first part is on the properties of ordinary Portland
cement (OPC), particularly, based on nanoindentation technique. The second part is a series
of an experiment on blended cement using statistical analysis tools. Ordinary Portland cement
(OPC) is made primarily from calcareous materials such as chalk or limestone, and argillaceous
materials such as shale or clay. This material is burnt at a high temperature of around 1450◦C
in a rotary kiln to form clinker which is then grounded with a requisite amount of gypsum into
a fine powder known as Portland cement. ASTM ASTM Standard C150 (2015) has designed
six types of Portland cement as Type I to V and White, differ primarily in aluminate (C3A)
contents and fineness. The general characteristics of different types of cement are listed in Table
4.1. Two important constitutes in cement are Alite (C2S) and Belite (C3S) (Neville, 2011).
In Australia, available cement types are General purpose cement (GP), General purpose lime-
stone cement (GL), Blended cement (GB), and Special purpose cements such as high early
strength cement (HE), low heat cement (LH), sulphate resisting cement (SR), and shrinkage
limited cement (SL) as shown in Table 4.2 (Australia Standard, 2010).
OPC is the main binder of concrete thus a large amount of OPC is produced due to the high
volume of concrete used worldwide (Flatt et al., 2012). It is, therefore, necessary, as a reference,
to know the properties of OPC in detail so it can be optimised. In this research, nanoindentation
technique is used for this purpose.
The first part of this Chapter, Section 4.2, presents indentation properties of OPC paste, and
the second part, Section 4.3, presents the mechanical properties of blended cement, paste and
mortar. An experimental programme of blended cement mixtures was designed according to
Taguchi’s approach (Roy, 2010), which is an efficient method for investigating optimum design
Table 4.1: General features of types of cement in ASTM C150 (Neville, 2011)
Type Classification Characteristics
I General purpose High C3S contentII Sulphate resistance Low C3A content (<8%)III High early strength Fine groundIV Low heat (slow reaction) Low C3S content (<50%) and C3AV High sulphate resistance Low C3A content (<5%)
White White colour Low MgO, No C4AF
47
Table 4.2: Properties and characteristics of cements in AS 3972-2010 (Australia Standard, 2010)
TypeSetting time Temperature Expansion Shrinkage
min max max. rise max. at 16 weeks max. 28 days(min) (hours) (◦C) (microstrain) (microstrain)
GP 45 6 - - -GL 45 10 - - -GB 45 10 - - -HE 45 6 - - -LH 45 10 23 - -SR 45 10 - 750 -SL 45 10 - - 750
parameters for the required performances. Also, Analysis of variance (ANOVA) and regression
were conducted to understand the better statistical relationship between properties on different
parameters and influence of parameters that contribute to measured variation of properties.
4.2 Indentation Properties of Ordinary Portland cement
Nanostructure of concrete is controlled by the structure of calcium silicate hydrated (CSH)
which governs fundamental properties such as strength, relaxation, creep, and fracture be-
haviour of hydrated cement. This demands a detailed knowledge of the nanostructure and how
it relates to local mechanical properties. Thus, it is important to understand nanostructure of
OPC. Several techniques have been used to understand CSH structure. Thomas et al. (1998)
studied determine two CSH morphologies using nitrogen but results were varying parameters
such as internal water content. Similarly, Tennis and Jennings (2000), revised model of the
microstructure of OPC that quantitatively predicts the volume of various phases. This model
provides a mean for quantifying the volumetric proportion of major hydration products such as
capillary porosity (MP), low-density CSH (LD-CSH) and high-density CSH (HD-CSH). Con-
stantinides and Ulm (2004) determined two types of CSH with Portlandite (CH), and clinker
with nanoindentation. The result shows that decalcification of CSH phases is the primary source
of nanometer-scale elastic modulus degradation. Generally, the complex viscous characteristic
of OPC is according to the viscous behaviour of CSH, which is the primary hydration product
of OPC. Advances in cement technology have enabled to understand the properties of CSH
recently. However, CSH exhibits significant local variations and is still difficult to probe into
its overall characteristics directly. Thus, a study on viscoelastic properties of hydration prod-
ucts by nanoindentation is well-suited, making it possible to probe sub-micrometric volumes of
material (Vandamme and Ulm, 2013).
In this section, therefore, it will present an investigation of indentation properties of hydration
products of OPC using deconvolution technique, micro poromechanics, and time-dependent
properties using nanoindentation as introduced in Chapter 3.
4.2.1 Sample Preparation
General purpose (Type I) Portland cement available locally in Australia was used to prepare
OPC paste sample for nanoindentation test. Three specimens were cast in 50mm cubic mould
with 0.3 of water to cement ratio, based on workability of mixture. Specimens were cured in
ambient curing at 23◦C ± 3. The compressive strength of specimens was obtained around 94.6
MPa. Each specimen was cut using a diamond saw after 28 days of curing to get the core
part of 10mm long segment. The hydration of specimens was arrested using solvent exchange
48
and oven dry method (Zhang and Scherer, 2011). Miller et al. (2008) pointed out that it was
important to reduce the surface roughness of specimen to get an accurate nanoindentation
results because the specimen’s surface has a significant influence on the test results (ISO 14577-
1, 2002). Therefore, fine emery paper was used to grind all specimens to reduce the surface
roughness. After that, the surface was also polished by using a suspension solution ranging from
6.0µm to 0.1µm for 45 mins and with 0.05µm for another 15mins; noting that the polishing
effect would last for around 8 hours. Nanoindentation test was then carried out on three sets
of 10mm cubic specimens with XP testing method; each set was specified 10 × 10 grid (100
indentations) with 20µm grid spacing, i.e., approximately 300 points were impressed on the
sample. Maximum load of Pmax =0.5mN applied in 15 seconds, kept constant 15 seconds and
unloaded in 10 seconds. The Berkovich indenter tip was used and Poisson’s ratio was assumed
0.25.
4.2.2 Statistical Indentation
The deconvolution technique presented in Section 3.5 can be used to analyse the mean properties
and volume fraction of each phase from a grid indentation. This process was carried out using
MATLAB (2014). The deconvolution technique process requires input parameters such as initial
values and the number of material phase. It has been well presented that hydration products
of OPC have five significant mechanical phases: capillary pores (MP), two CSH, which are
low-density (LD) CSH and high density (HD) CSH, Portlandite (CH), and unhydrated clinker
(Constantinides and Ulm, 2007; Velez et al., 2001).
The indentation modulus (M) and hardness (H) are determined using five Gaussian deconvolu-
tion input parameters as presented in Section 3.2. The results of deconvolution are illustrated
in Figure 4.1 and given in Table 4.3. The mean value of the first peak shows M = 9.388
GPa , H = 0.297 GPa; the second peak, M = 16.591 GPa, H = 0.7 GPa, which are in good
agreement with literatures (Constantinides and Ulm, 2007; Velez et al., 2001). For the third
peak, M = 29.818 GPa, H = 1.412 GPa; the fourth peak M = 48.677 GPa, H = 10.495 GPa;
and the clinker, M = 112.378 GPa, H = 14.718 GPa. Therefore, it can be identified that
the hydration products of OPC are made of five mechanical phases and a deconvolution gird
indentation can be performed with five Gaussian deconvolution input parameters. The typical
indentation load-depth (P − h) curves of indented material are presented in Figure 4.2.
As mentioned in Section 3.4, with the hypothesis that OPC is a porous material, the quadratic
minimisation problem, Equation (3.4.8), can be solved. The quadratic minimisation procedure
of scaling indentation modulus and hardness exhibit the properties of the hydration products.
The packing density relationship distribution can be plotted by experimental and model values
as shown in Figure 4.3. The relative error between linear, Equation (3.4.6), and non-linear,
Equation (3.4.5), the scaling model and the experimental values of the minimisation are: eM =
Table 4.3: Deconvolution results for indentation modulus and hardness
PhaseM H f
mean StD mean StD %
MP 9.388 3.562 0.297 0.116 43LD CSH 16.591 4.712 0.700 0.147 34HD CSH 29.818 3.984 1.412 0.215 10
CH 48.677 10.495 2.352 1.239 8Clinker 112.378 29.557 14.718 8.027 5
49
Figure 4.1: Deconvolution technique of indentation modulus and hardness
Figure 4.2: Typical indentation load-depth (P − h) curves
50
14.042 ± 13.215 GPa for indentation modulus and eH = 9.681 ± 9.065 GPa for indentation
hardness. The solid properties of hydration products are: stiffness (λ) = 189.442 GPa, Poisson’s
ratio (vs) = 0.499, cohesion (κ) = 1.300 GPa, friction coefficient (ζ) = 0.566 and the friction
angle is calculated from friction coefficient (ζ) as θ = 29.544 degree. The solid properties
of hydration products can be understanding in the sense of the Druker-Parger strength and
Coulomb material models (Ulm et al., 2007). The deconvolution technique using Equation
(3.5.3), the indentation modulus (M), hardness (H), packing density (η) and volume fraction
(f) are obtained. The details are presented in Table 4.4 and the deconvolution results of packing
density are given in Figure 4.4.
CSH is a major hydration product which relates directly to mechanical strength (Berry et al.,
Figure 4.3: Packing density (η) distribution
Table 4.4: Deconvolution results for indentation modulus and hardness
PhaseIndentation Modulus Indentation Hardenss Packing density Volume fraction
M H η f
mean StD mean StD mean StD %
MP 9.467 3.604 0.298 0.118 0.531 0.01 44LD CSH 16.787 4.804 0.704 0.144 0.556 0.01 33HD CSH 30.481 4.257 1.415 0.222 0.595 0.01 10
CH 51.449 9.831 2.579 1.170 0.627 0.06 7Clinker 114.539 31.598 14.387 7.221 0.780 0.12 5
Figure 4.4: Deconvolution result of packing density
51
1990; Neville, 2011). The lowest indentation modulus, hardness and packing density which can
be observed are associated with a capillary pores (MP) (Constantinides and Ulm, 2007). In the
present case, it was found that pores phase has the volume proportion of 44% of the total volume
fraction of the overall hydration products, and has the indentation modulus M = 9.467± 3.604
GPa, hardness H = 0.298 ± 0.118 GPa, and packing density η = 0.531 ± 0.01. LD CSH
phase has a characteristic mean stiffness of indentation modulus M = 16.787 ± 4.804 GPa,
hardness H = 0.704 ± 0.144 GPa, and packing density η = 0.556 ± 0.01, with 33% of volume
fraction. The stiffness of indentation modulus, hardness, packing density and volume fraction
of HD CSH are: M = 30.481 ± 2.441 GPa, H = 1.415 ± 0.222 GPa, and packing density
η = 0.595±0.01, with 10% of volume fraction. The Portlandite (CH) phase has the indentation
modulus M = 51.449 ± 9.831 GPa, hardness H = 2.579 ± 1.170 GPa, and packing density
η = 0.627± 0.06 with 7% of the total volume fraction. The indentation modulus, hardness and
packing density of clinker are M = 114.539 ± 31.598 GPa, hardness H = 14.387 ± 7.221 GPa,
and packing density η = 0.780 ± 0.12 with 5% of volume fraction. Application of Equation
(3.4.9) to the indented test data yields total porosity (ϕ) = 0.357. The statistical indentation
technique provides a new nonintrusive way of determining the porosity of nanogranular material
(Ulm et al., 2007).
The deconvolution technique results proved the mechanical properties at a point of the inden-
tation grid. The information of these nodal values provides a morphological arrangement of the
different phases in the composite material (Constantinides et al., 2006). The contour map of the
indentation modulus of the hydration phases (MP, LD CSH, HD CSH, and CH) was constructed
using MATLAB (2014) as illustrated in Figure 4.5. The information on the contour map of the
mechanical properties provides the microstructural distribution of each phase in the specimen
with respect to the x-y plane. Showing localised effect in a small area is a good technique to
display the volume fraction of different phases. The contour map shows that higher modulus
is surrounding by lower modulus, i.e., HD CSH phase is surrounding by LD CSH phase, CH
phase is surrounding by HD CSH phase, and unhydrated phase (clinker) is surrounding by CH
phase.
Figure 4.5: Contour map of hydration products of OPC paste. Image size is 180µm × 180µm with20µm grid spacing
52
4.2.3 Viscoelastic Properties
A classical principle of elastic properties deals with the mechanical properties of elastic solids in
accordance with Hooke’s law, i.e., stress is directly proportional to strain in small deformation
and independent of the rate of strain (Ferry, 1980). Material behaviour exhibiting viscoelastic
characteristics has ability to store energies under load-deformation upon removal of stress and
return to its original state. The measurement of viscoelastic properties of the material can be
obtained considering the nature and the rate of configurational rearrangements and interaction
of properties. Nanoindentation is a new technique that can determine quantitative viscoelastic
properties of the tested material. The time-dependent properties of the material are generally
obtained in term of relaxation and creep properties using indentation results.
4.2.3.1 Contact Relaxation Modulus
Classical linear viscoelastic properties, also known as rheological models, such as Maxwell model,
three-element Kelvin-Voigt model and combined four-element Maxwell-Kelvin-Voigt model can
be obtained from indentation tests. Rheological models are associated with a part on the
nanoindentation load-displacement (P − h) curves, i.e., in the holding period at the maximum
load. In the present case, the P − h curves from the nanoindentation tests were fitted to the
Equations (3.8.20) to (3.8.22) using a non-linear least squares method in MATLAB (2014).
Based on the results, the corrected coefficients of Maxwell, Kelvin-Voigt and Maxwell-Kelvin-
Voigt model were observed as 0.902±0.046, 0.936±0.242 and 0.987±0.028, respectively. Table
4.5 summarises the Rheological properties of OPC. The results show that standard deviation
is significant because of results included total hydration products of Rheological element.
The relaxation modulus is greatly significant for engineering applications. The normalised
contact relaxation modulus, which is defined as the relaxation modulus divided by the instan-
taneous modulus represents important material properties of the linear viscoelastic material
(Cao et al., 2010). The mean normalised contact relaxation modulus is shown in Figure 4.6.
The contact relaxation modulus based on Maxwell model shows a decrease in normalised re-
laxation modulus because Maxwell model is a linear function. The results of Kelvin-Voigt and
Combined Maxwell-Kevin-Voigt of contact normalised relaxation modulus indicate that the
initial reduction of relaxation modulus is small (around 3.5% in 10 second). Also, normalised
relaxation modulus can be represented by the dynamic modulus, as discussed in Chapter 3.
The results indicate that initial reduction of elastic modulus while constant strain. It means
that the stress relaxation and reduction of elastic modulus can be negligible during early curing
ages of OPC pastes.
4.2.3.2 Contact Creep Compliance
The contact creep compliance Jc was introduced in Section 3.8. Generally, the complex creep
behaviour of cementitious material is largely related to the viscoelastic response of the vital
hydration or reaction products and the binding phase of hardening. The creep compliance
depends on the geometry of the axisymmetric indenter and maximum value of control variables
Table 4.5: Rheological properties
Maxwell Kelvin-Voigt Maxwell-Kelvin-Voigt
Mo (GPa) 14.532± 33.186 14.977± 34.347 15.074± 34.662Mv (GPa) - 143.412± 356.395 254.983± 574.723ηM (GPa s) 1795.889± 4497.288 - 3475.169± 8979.073ηv (GPa) - 566.122± 1192.053 427.100± 835.500
53
Figure 4.6: Normalised relaxation modulus
in indentation process (Neville, 2011).
For the long-term contact creep compliance, the change in depth of the creep phase was fit with
a logarithmic function in Equation (3.8.33). The curve fitting was performed by MATLAB
(2014). The average corrected coefficient of the fitted curve was observed as 0.986± 0.04. The
typical curve fitting in penetration depth versus time is presented in Figure 4.7. The average
error of x1 coefficient was captured as 4.858± 2.455 nm.
With the logarithm curve fitting results, x1, the long-term contact creep compliance rate can
be obtained as:
Jc =1
Ctwhere C =
πPmax4 tan (a)hmaxx1
(4.2.1)
Figure 4.7: Typical logarithm curve fitting in creep phase
54
where C is contact compliance creep modulus. As explained in Section 3.8, Pmax, hmax and
α are obtained from indentation test and geometry properties. The long-term contact creep
modulus was determined as 409.266 GPa. Figure 4.8 shows the average long-term contact
creep compliance rate. The result of creep compliance rate clearly shows that after a few days
of applying stress, the rate of the creep compliance sharply decreases. The results of creep
behaviour are useful in quantifying a unique mechanical response of time-dependent materials.
From the results of contact creep compliance rate of OPC, specific creep after one year was
determined by 18.32 microstrain/MPa.
As presented earlier in Section 4.2.2, the deconvolution technique identified material phases as
MP, LD CSH, HD CSH, CH and un-hydrated clinker. In addition, this deconvolution technique
can identify the contact creep modulus C of each phase. The greater the C value represents the
lower rate of the creep. Table 4.6 summarises the results of M , H, C and η with 5 Gaussian
deconvolution input parameters. For MP phase: C = 74.552 ± 37.754 GPa; LD CSH phase:
C = 162.201 ± 53.582 GPa; HD CSH Phase: C = 333.797 ± 46.979 GPa; and CH phase:
C = 663.387±188.444 GPa. The CDF and PDF results of contact creep modulus are presented
in Figure 4.9. Figure 4.10 shows a comparison between contact creep compliance rates of all
the hydration phase. MP phase has the highest creep compliance rate of OPC because it has
a highest creep modulus. It means that when the porosity increases, it tends to increase the
creep compliance.
4.2.4 Indentation Stress-Strain Curve and Fracture Toughness
The relationship between the maximum depth h and the contact depth, hc = f (h), depends on
the indenter tip geometry and indented material. The function, hc = f (h), can be measured by
contact stiffness measurement (CSM) between the indented material and the indenter process
using Equation (3.9.2). Knowing hc = f (h) function, and the geometry of indenter tip, the
instant contact area can be determined by Equation (3.9.1) which is the instant contact area
with respect to the instant stress for loading process. Using Equations (3.9.3) to (3.9.5), the
indentation stress-strain curve can be obtained as presented in Section 4.2. The results of
the indentation stress-strain curves of statistical deconvolution phases show that each phase
Figure 4.8: Contact creep compliance rate
55
Table 4.6: Deconvolution results of indentation modulus, hardness, contact creep modulus andpacking density
PhaseIndentation Indentation Contact creep Packing Volume
Modulus Hardness Modulus Density fractionM H C η f
mean StD mean StD mean StD mean StD %
MP 9.292 3.509 0.298 0.114 75.552 37.754 0.531 0.011 42LD CSH 16.326 4.510 0.696 0.146 162.201 53.582 0.556 0.014 33HD CSH 29.028 4.056 1.391 0.232 333.797 46.979 0.585 0.041 10
CH 45.755 11.384 2.107 1.261 663.387 184.444 0.599 0.028 8Clinker 106.234 34.349 13.160 8.810 3948.108 1388.759 0.783 0.088 6
Table 4.7: Deconvolution results of indentation modulus, hardness, contact creep modulus andpacking density
PhaseIndentation Indentation Contact creep Packing Volume
Modulus Hardness Modulus Density fractionM H C η f
mean StD mean StD mean StD mean StD %
MP 9.292 3.509 0.298 0.114 75.552 37.754 0.531 0.011 42LD CSH 16.326 4.510 0.696 0.146 162.201 53.582 0.556 0.014 33HD CSH 29.028 4.056 1.391 0.232 333.797 46.979 0.585 0.041 10
CH 45.755 11.384 2.107 1.261 663.387 184.444 0.599 0.028 8Clinker 106.234 34.349 13.160 8.810 3948.108 1388.759 0.783 0.088 6
Figure 4.9: Deconvolution result of contact creep modulus
56
Figure 4.10: Contact creep compliance rate of hydration phases
can be captured based on initial and stationary yielding using Berkovich indenter tip (20nm
of the radius). The indentation stress-strain curves clearly illustrate elastic, elastic-plastic and
plastic regimes. The initial yielding points are observed approximately when the modulus values
are 70 MPa, 79 MPa, 200 MPa, and 425 MPa for MP, LD CSH, HD CSH, and CH phases,
respectively. The stationary yielding points from indentation stress-strain curves are obtained
when the strength values reach 354 MPa, 420 MPa, 580 MPa, and 1200 MPa for MP, LD CSH,
HD CSH, and CH phases, respectively. Based on the indentation stress-strain curves, MP phase
has small strain capacity in the fully plastic regime, as seen in Figure 4.11(a). The indentation
stress-strain curves of LD CSH and HD CSH phases, Figure 4.11(b) and (c), illustrate that
indentation strain hardening increases in the fully plastic regime and fracture point cannot
be observed. Figure 4.11(d) presents the indentation stress-strain curve of CH phase which
shows that it remains stable in the fully plastic regime. With the indentation stress-strain
curve of each phase of OPC paste, it can be explained that the strength failure occurred in MP
phase because of the small strain capacity compared to other phases. Therefore, high capillary
porosity will lead to low compressive strength. Chindaprasirt et al. (2005) reported that fine
fly ash (grinded) contents in mixture resulted in low capillary porosity and higher compressive
strength than the original fly ash (not grinded) contents in the mixture. Also, an increase in
sand to cementitious material ratio can increase capillary porosity and lead to low compressive
strength. The present finding validates theses points.
The fracture toughness of a material is linked to the energy stored in the form of elastic or
plastic energy before fracturing. As presented in Section 3.6, a robust method for determining
fracture toughness using Equations (3.6.3) to (3.6.9) was proposed. The overall indentation
results of the fracture energy release rate (Gc) and fracture toughness (Kc) were obtained as
0.526 N/m2 and 0.368 MPa m1/2, respectively. Shah (1989) determined fracture toughens of
cement pastes by conventional test method (dye penetration technique) using varying water to
cementitious material ratio. The obtained results in the present case are within the range of the
fracture toughness of cement pastes determined Shah (1989), which are 0.3 to 0.5 MPa m1/2.
57
Figure 4.11: Indentation stress-strain curves (a) MP (b) LD CSH (c) HD CSH (d) CH
4.2.5 Relationship between Indentation Properties
This section presents the results to illustrate the relationship between the indentation properties,
which are indentation modulus, hardness, packing density and volume fraction, as shown in
Figure 4.12. The indentation modulus and hardness deconvoluted data align well along the
curve. The fitting of power function to the indentation modulus and hardness relation yields
M = 22.83H0.904 with corrected coefficient 0.991 (99.1%). It was found that the relationship
between indentation modulus and hardness is almost linear, i.e., an increase in modulus leads
to an increase in the hardness. The relationship between indentation modulus and packing
density has fitting power function M = 1956(η − 0.5)1.659. The corrected coefficient is 0.971
(97.1%). This relationship shows that an increase in indentation modulus leads to an increase in
packing density. The deconvolution data of indentation modulus and contact creep modulus has
a relationship as M = 0.4357C0.7174 with corrected coefficient 0.992 (99.2%). In addition, the
contact creep modulus in an exclusive manner with the packing density. The fitting with power
function to the deconvoluted data yields C = 709224(η − 0.5)3.0463 with corrected coefficient
0.926 (92.6%). It was found that an increase in indentation modulus leads to a decrease in
creep modulus. The results of the relationship between indentation properties show power
relationship of the indentation modulus and creep modulus have a unique manner with the
packing density. An increase in packing density provides an increase in indentation modulus
and creep modulus. Moreover, the result of an increase in modulus and creep modulus delivers
an increase in hardness which provides high strength; and a decrease in creep compliance offers
low creep behaviour. Thus, the relationship between properties can be adapted for use as
a guideline in the design of OPC mixtures, i.e., an increase in packing density is the most
important objective in design of OPC mixtures became an increase in packing density provides
high strength and low creep behaviour in OPC mixtures.
58
59
Figure 4.12: Deconvoluted indentation properties with Power fit
60
4.3 Statistical Analysis of Properties of Blended Cement
Blended cement is a hydraulic cement that contains general purpose cement and pozzolan such
as fly ash and ground granulated iron blast-furnace slag (GGBF) (Australia Standard, 2010).
Blended cement has been used because of its low cost and improved properties. Fly ash is
commonly used blended cement that the main improvement to the strength of blended cement
is due to reaction the between calcium hydroxide (Ca(OH)2) which is a by-product of hydration
of OPC. When blended cement mixes with water, fly ash reacts with water and Ca(OH)2 to
form Calcium Silicate Hydrate (CSH) (Guide to the use of fly ash in concrete in Australia, 2009;
Shi, 1996). This section presents a study on blended cement mixtures using statistical analysis
such as Taguchi’s Design of Experiment, Analysis of Variance (ANOVA), and regression to
investigate statistical relationship properties of blended cement with different parameters.
4.3.1 Experimental Program
Taguchi’s design experimental approach to parameters design provides design engineers with a
systematic and efficient method for investigating optimum design parameters for performance
and cost. In this research, the following parameters are considered in the mix proportions.
- Fly ash (FA)
- Sand to cementitious material ratio (s/c)
- Water to cementitious material ratio (w/c)
- Superplasticiser (SP)
Four parameters and three levels of test variable were selected as shown in Table 4.8. The
Standard L9 (34) orthogonal array (Ross, 1996) according to the selected parameters was used,
as presented in Table 4.9.
Table 4.8: Variation parameters and levels
Levels FA s/c w/c SP
1 0 % 0 0.3 0 %2 10% 1.5 0.35 0.1%3 20% 2.5 0.4 0.2%
Table 4.9: Standard L9 orthogonal array
Mix FA s/c w/c SP
1 1 1 1 12 1 2 2 23 1 3 3 34 2 1 2 35 2 2 3 16 2 3 1 27 3 1 3 28 3 2 1 39 3 3 2 1
In this experimental work, general purpose (Type I) Portland cement, and Class F (low calcium)
fly ash from Australia were used to prepare specimens. Table 4.10 shows the typical chemical
compositions of OPC and fly ash. The Loss on Ignition (LOI) of fly ash for the experiment
was 1.53% and the median particle size was 45µm. The sand (SSD: saturated surface dried
61
condition) with a specific gravity of 2.6 was used as fine aggregate. The XRD pattern of OPC
and fly ash shown in Figure 4.13. The specimens were cast in 50mm cubic mould. Blend cement
mixtures were cured in ambient curing at 23◦C ± 3 until testing. The compressive strength of
the specimens were tested at a loading rate of 900N/s with a universal testing machine at the
age of 28 days in accordance with ASTM ASTM Standard C109 (2011). Also, the density of
the specimen was measured at the age of 28 days in accordance with ASTM ASTM Standard
C642 (2013). A naphthalene sulphonate superplasticiser was used to improve workability. The
reported results are the average of five samples.
Table 4.10: Chemical composition of OPC (Type I) and low calcium fly ash (wt. %)
SiO2 Al2O3 CaO Fe2O3 K2O MgO SO3
OPC 21.1 4.7 63.6 2.7 - 2.6 2.5Fly ash 65.9 24.0 1.59 2.87 1.44 - -
Figure 4.13: XRD pattern of (a) OPC (Type I) and (b) low calcium fly ash
62
4.3.2 Density
Density is important in reducing the self-weight. Figure 4.14 and Table 4.11 show the average
density results of the tested blended cement mixtures. The results indicate that most of the
mixtures have fairly uniform density during the 7 to 28 days of curing ages. The paste samples
with fly ash, Mixes 4 and 7, have the lowest density of 1929 kg/m3 and 1776 kg/m3, respectively.
The mortar samples without fly ash, Mixes 2 and 3, have the highest density of 2164 kg/m3
and 2145 kg/m3, respectively. Overall, the density of all the mixtures slightly decreases after
curing for 28 days, with an exception that the mortar samples, Mixes 8 and 9, which contained
20% of fly ash to cement ratio, have a slight increase in the density during 7 to 14 days of
curing age. Based on these results, a statistical S/N ratio analysis was performed to determine
the effect of these parameters on the density of the mixtures, as shown in Figure 4.15, which
can be seen that sand to cementitious material ratio has the most effect on the density at 28
days of curing age. ANOVA was performed on the results from the nine mixtures. ANOVA
results indicate that sand to cementitious material ratio contributed 54% to the density at 28
days of curing age as presented in Table 4.12 and Figure 4.16. It was observed that increasing
fly ash contents resulted in decreasing the density. The contribution of fly ash on the density is
determined as 24%. The contribution of increasing the water to cementitious material ratio on
the density is 14%. Superplasticiser has a minor influence on the density at 28 days of curing
age; its contribution is obtained as 5%. As expected, the sand to cementitious material ratio
was the most important parameter on density.
Figure 4.14: Density with curing age
63
Table 4.11: Density with curing age results (kg/m3)
Mix 7 days 14 days 28 days
1 2082 2073 20702 2164 2153 21463 2145 2132 21264 1929 1934 19275 2056 2030 20276 2115 2111 21107 1776 1773 17628 2090 2111 20779 2118 2126 2091
Figure 4.15: Density with curing age
Table 4.12: ANOVA results on density
Source DFa SSb MSc Contribution %
7 days density
FA 2 29447 14724 23.84s/c 2 69882 34941 56.57w/c 2 17432 8716 14.11SP 2 6773 3386 5.48
14 days density
FA 2 22893 11477 18.86s/c 2 68161 34081 56.15w/c 2 23721 11861 19.54SP 2 6626 3313 5.49
28 days density
FA 2 29462 14731 24.82s/c 2 63513 31756 53.51w/c 2 20772 10386 17.50SP 2 4954 2477 4.17
adegree of freedom bsum of square cmean square
64
Figure 4.16: Contribution of experimental parameters on density
65
4.3.3 Compressive Strength
The compressive strength results are presented in Table 4.13 and Figure 4.17(a). The average
percentage increase in compressive strength at 7 days of curing age is 8% of 28 days compressive
strength. The paste mixture Mix 1 has the minimum compressive strength 87.8 and 94.5 MPa
at 7 days and 28 days of curing age, respectively. The mortar mixture Mix 5, with 10% of
fly ash, 0.15 of sand to cementitious material ratio, 0.4 of water to cementitious material ratio
and no superplasticiser has the minimum compressive strength of 43.6 MPa and 54.5 MPa at
7 days and 28 days of curing age, respectively. Figure 4.17(b) shows the compressive strength
development between 7 days to 28 days of curing age. Mix 4 has the highest compressive
strength gain between 7 to 28 days of curing age with 21.62 MPa. The compressive strength
gain in Mix 6 between 7 to 28 days is the lowest, which is 4.31 MPa. Therefore, Mix 5 has
minimum compressive strength but the highest compressive strength gains between 7 to 28 days
of curing ages.
Table 4.13: The results in based on compressive strength (MPa)
Mix7 days 14 days 28 days
fck,cubic StD fck,cubic StD fck,cubic StD
1 87.82 2.44 91.37 1.74 94.56 2.262 63.54 1.56 67.42 3.52 72.82 4.353 53.22 1.61 58.78 2.64 62.63 2.864 59.59 5.68 73.34 4.65 81.20 3.455 43.65 3.44 51.61 2.40 54.55 2.796 52.01 1.36 55.22 2.27 56.32 1.847 51.39 3.16 58.62 2.62 58.79 3.068 54.76 2.92 65.05 4.11 67.14 6.999 47.61 2.09 58.44 2.15 59.53 3.51
66
Figure 4.17: (a) compressive strength development and (b) % of compressive strength developmentbetween 7 to 28 days
67
The typical XRD pattern of OPC paste (Mix 1) is as presented in Figure 4.18. Highly notice-
able peaks are “Calcium Silicate Hydrated (CSH)”, which is main hydration products of the
cementing compound, “Portlandite (CH)” and “Dicalcium Silicate (C2S)”. Due to curing ages,
XRD pattern indicates that CSH phase increases. Scanning Electron Microscope (SEM) also
shows the growth of hydration products as fabric structures when curing proceeds as presented
in Figure 4.19.
Figure 4.18: XRD pattern of Mix 1
Figure 4.19: SEM image of Mix 1
68
The S/N ratio graphs of parameters for compressive strength at 7 days and 28 days of curing
age are shown in Figure 4.20. Fly ash has the most significant effect on 7 days compressive
strength. On 28 days compressive strength, however, sand to cementitious material ratio is the
most affecting parameter. ANOVA results show sand to cementitious material ratio has 41.56%
contribution on 7 days and 39.59% contribution on 28 days, as shown in Table 4.14 and Figure
4.21. The effect of superplasticiser on 7 days and 28 days compressive strength is 2.32% and
7.64% of contribution, respectively. The S/N ratio of fly ash and sand to cementitious material
ratio at 7 days and 28 days of curing age results show similar patterns. The S/N ratio of water
to cementitious material ratio shows that water to cementitious material ratio of 0.3 to 0.35 has
not significant effect on 28 days compressive strength. ANOVA results of sample containing
water to cementitious material ratio present mostly similar contributions of curing ages which
are in range 25.32% and 27.49%. Therefore, an increase in fly ash, sand to cementitious material
ratio and water to cementitious material ratio in blended cement mixtures lead to a decrease
in the compressive strength.
Table 4.14: ANOVA results on compressive strength
Source DFa SSb MSc Contribution %
7 dayscompressivestrength
FA 2 557.51 278.75 41.56s/c 2 394.68 197.34 29.42w/c 2 358.14 179.07 26.70SP 2 31.18 15.59 2.32
14 dayscompressivestrength
FA 2 295.47 474.02 25.36s/c 2 474.02 237.01 40.69w/c 2 3 20.24 160.12 27.49SP 2 75.28 37.64 6.46
28 dayscompressivestrength
FA 2 385.4 192.69 27.45s/c 2 555.9 277.95 39.59w/c 2 355.6 177.78 25.32SP 2 107.3 53.67 7.64
adegree of freedom bsum of square cmean square
69
Figure 4.20: Effect of parameters on compressive strength
70
Figure 4.21: Contribution of experimental parameters on compressive strength
71
The results of S/N on compressive strength gain between 7 days to 28 days of curing ages
indicate that superplasticiser is the most significant parameter as shown in Figure 4.22. Sand
to cementitious material ratio has a minor effect on the compressive strength gain. ANOVA
results reveal that the contribution of superplasticiser on compressive strength gain is 44.01%
and sand to cementitious material ratio is 9.19%, as shown in Table 4.15. The optimal mixing
design for compressive strength gain on blended cement mixtures is 20% of fly ash, 1.5 of
sand to cementitious material ratio, 0.35 of water to cementitious material ratio and 0.2% of
superplasticiser.
Figure 4.22: Effect of parameters on compressive strength gain
Table 4.15: ANOVA results on density
Source DFa SSb MSc Contribution %
Compressivestrengthgain
FA 2 21.7 10.85 11.19s/c 2 17.83 8.915 9.19w/c 2 69.09 34.546 35.61SP 2 85.37 42.687 44.01
adegree of freedom bsum of square cmean square
72
Figure 4.23: Contribution of experimental parameters on compressive strength gain
4.3.4 Water Absorption
Low permeability, especially resistance to freezing and thawing is one of the most important
properties of the good quality of cement binder. The permeability of cement binder differs
from absorption and relates to the size of pores, its distribution, and most importantly its
continuity. Therefore, good quality of binder cement has low water absorption and high density
(Neville, 2011). According to ASTM ASTM Standard C140 (2015), water absorption rate of
each specimen was determined. The results of water absorption test indicate that Mix 7 has
the highest water absorption, which is 18.02%. Mix 3 has the lowest water absorption rate of
7.45%, as shown in Figure 4.24.
Figure 4.24: Water absorption results
73
The results of S/N ratio on water absorption show that sand to cementitious material ratio
has a major effect on the water absorption, as shown in Figure 4.25. The increase of sand to
cementitious material ratio leads to the decrease of water absorption on blended cement mixture.
Superplasticiser is a minor important parameter on water absorption. ANOVA results show
that the contribution of sand to cementitious material ratio is 85.44% and superplasticiser is
1.92%, as shown in Table 4.16. Thus, an increase of fine solid such as fly ash and OPC could
lead to increase in water absorption. Moreover, high contents of superplasticiser also decrease
the water absorption. The optimised mixing proportion for water absorption is: no fly ash, 2.5
of sand to cementitious material ratio, 0.3 of water to cementitious material ratio and 0.2% of
superplasticiser.
Figure 4.25: Effect of parameters on water absorption results
Table 4.16: ANOVA results on density
Source DFa SSb MSc Contribution %
Waterabsorption
FA 2 5.717 2.859 5.15s/c 2 94.8 47.4 85.44w/c 2 8.312 4.156 7.49SP 2 2.126 1.063 1.92
adegree of freedom bsum of square cmean square
74
Figure 4.26: Contribution of experimental parameters on water absorption
4.3.5 Regression Analysis
Regression analysis is a statistical tool for estimation of relationships between variables. One
of important reasons to determine the regression model is to uncover causes by studying the
relationship between variables. Sometimes, statistical relationship does not necessarily imply
causal relationship but the presence of relationship can give a good starting point for research.
If statistical confidence is indicated with the regression model, then values of the explanatory
variables can be used predict the output variables. Another reason for regression analysis is
to examine the test hypotheses (Seber and Lee, 2012). In this research, therefore, regression
analysis was used as a tool for estimating the relationship between variables and to predict
outside of parameters ranges of explanatory variables.
Through linear regression analysis, the empirical relationship of blended cement mixtures at 28
days curing ages was obtained as:
ρ28 = 1747 + 5478.9x1 + 44.308x2 + 1086.8x3 − 43947x4
+ 209.85x1x2 − 18486x1x3 − 73.835x2x3 (4.3.1a)
fck,cubic,28 = 199.11− 745.64x1 − 49.188x2 − 347.35x3 + 9736.9x4
+ 79.579x1x2 + 1716.3x1x3 + 106.24x2x3 (4.3.1b)
where x1, x2, x3, x4 denote fly ash, sand to cementitious material ratio, water to cementitious
material ratio and superplasticiser, respectively. As shown in Table 4.17, the empirical equation
is in good agreement with the experimental results.
Validation of predictive model can be carried out with corrected coefficient (Analla, 1998). The
information about regression analysis is presented in Table 4.18. It can be seen that corrected
coefficient values of density and compressive at 28 days are 0.980 (98%) and 0.983 (98.3%),
respectively. These corrected coefficients indicate that empirical model states the good fit of
the model being validated.
Several researchers (Ozbay et al., 2009; Srinivasan et al., 2003; Tanyildizi, 2013; Xu et al., 2012)
reported successful application of Taguchi’s design of experiment with ANOVA technique to
75
Table 4.17: Experiment and Predicted density (kg/m3) and compressive strength (MPa) at 28 days
MixExperiment Predicted Ratio
ρexp fck,cubic,exp ρpred fck,cubic,predρexp fck,cubic,exp
/ρpred /fck,cubic,pred
1 2070 94.56 2073 94.90 0.998 0.9962 2146 72.82 2112 69.40 1.016 1.0493 2126 62.63 2131 63.14 0.998 0.9924 1927 81.20 1941 82.57 0.993 0.9835 2027 54.55 2044 56.26 0.992 0.9706 2110 56.32 2131 58.37 0.990 0.9657 1762 58.79 1755 58.11 1.004 1.0128 2077 67.14 2068 66.28 1.004 1.0139 2091 59.53 2081 58.50 1.005 1.018
Average 1.000 1.000StD 0.008 0.026
Table 4.18: Regression analysis of test data
(a) Density at 28 days
Parameters Estimate Standard error T value
Intercept 1747.4 831.2 2.102FA 5748.9 5737.1 0.995s/c 44.308 270.45 0.164w/c 1086.8 2697.2 0.403SP -43947 62285 -0.706
FA×s/c 209.85 614.53 0.341FA×w/c -18486 16038 -1.523s/c×SP -73.835 909.15 -0.081
Corrected coefficient = 0.980
(b) Compressive strength at 28 days
Parameters Estimate Standard error T value
Intercept 199.11 82.509 2.413FA -745.64 569.49 -1.309s/c -49.118 26.846 -1.829w/c -347.35 267.73 -1.297SP 9763.9 6182.7 1.579
FA×s/c 79.579 61.002 1.305FA×w/c 1716.3 1592.1 1.078s/c×SP 106.24 90.247 1.177
Corrected coefficient = 0.983
76
cementitious-based binders. In this research, Taguchi’s design of experiment results are in good
agreement with the results in literature (Bye, 1999; Neville, 2011; Peiwei et al., 2001; Siddique,
2004), the increase in fly ash, sand to cementitious material ratio, water to cementitious ma-
terial ratio reduces the compressive strength. Thus, Taguchi’s design of experiment leads to
identifying the optimal mixing design with S/N ratio. ANOVA technique helps to determine the
parameters contribution in a measured variation of properties that indicate most considerable
parameters on mixing design.
4.3.6 High Temperature Exposure
Concrete can expose to high temperature during a fire or when it is near to furnace and
power reactions. During its exposure to high temperatures, the mechanical properties, such
as strength and elastic modulus, deteriorate and decrease significantly (Morsy et al., 2009).
Physical deterioration process affects the durability of the concrete structure, therefore, is an
important issue during and after high temperature exposure. The harmful effects of high
temperature on concrete can be minimised by taking a preventive measure such as choosing the
right material (Aydın, 2008). OPC is a common binder in concrete. However, when exposed
to high temperatures, spalling of OPC-binder concrete could occur, resulting in a rapid loss
of concrete cover, layer by layer, potentially exposing the main steel reinforcement. Therefore,
it is beneficial to improve the property of cement binder that carries good resistance to high
temperature in term of spalling resistance and strength loss (Kong and Sanjayan, 2010).
Chan et al. (1996) studied high strength and normal strength concrete exposed to high tem-
peratures in a range between 400 to 1200◦C. They found that temperature between 400◦C
and 800◦C was a critical range of a loss of concrete strength. The normal strength concrete
loses around 10 to 25% and 80% of its original compressive strength at 400◦C and 800◦C, re-
spectively. Many research papers reported that using pozzolan as partial cement replacement
could lead to an improvement in temperature resistant properties. Xu et al. (2001) studied the
influence of high temperature on fly ash in which OPC was replaced with 0%, 25% and 55%
of fly ash. The residual strength of concrete samples after exposing to 250◦C to 800◦C were
determined and found that a dosage of 55% fly ash in concrete led to high residual strength
than other dosages fly ash samples. Poon et al. (2001) conducted research to investigate the
effect of the residual strength and durability on normal and high pozzolanic concrete exposed to
high temperatures. They concluded that pozzolanic concrete has better performance than OPC
concrete after exposed to high temperatures. The mix containing 30% of fly ash replacement
has maximum relative residual strength. Also, most of spalling occurred between 400◦C and
600◦C but no spalling was observed in fly ash concrete. Mendes et al. (2008) used slag at 35%,
50% and 65% and determined the residual strength after exposing to 100◦C to 800◦C. It was
found that the samples with slag showed less damage under elevated temperatures. Although
numerous researchers studied the effect of high temperatures exposure on the properties of
blended cement mixtures, there is not enough on parameters contribution.
High temperature exposure test was conducted on selected five specimens of each mixtures as
shown in Table 4.9, after 28 days curing then dried in an oven at 105 ± 5◦C for 24 hours. After
drying, the specimens were placed in an electric kiln to be heated up to reaching maximum
temperatures of either 200, 400, 600 or 800◦C with 2 hours holding time. The heating rate
was 10◦C per minute. After exposed to temperature, the specimens were allowed to cool down
gradually to room temperature. Table 4.19 and Figure 4.27 present the residual strength of
blended cement mixture after exposure to high temperatures. The test results show that each
temperature range has a distinct pattern of strength gain or loss. It is observed that there is
77
a significant increase in strength around 18.3 MPa when mixtures are exposed to 200◦C. This
increase could be due to the hydration of unhydrated particles which were activated as a results
of temperature rise (Morsy et al., 2008). The maximum and minimum residual strength of Mix
1 and Mix 5 are 112.8 and 56.8 MPa, respectively, when exposed to 200◦C. With Mix 1, a
significant decrease in the strength is obtained when exposed to between 600 to 800◦C.
Table 4.19: Residual strength of blended cement mixtures (Unit in MPa)
MIX23◦ 200◦ 400◦ 600◦ 800◦
fck,cubic StD fck,cubic StD fck,cubic StD fck,cubic StD fck,cubic StD
1 94.5 2.2 112.8 5.7 97.5 3.3 57.0 4.9 - -2 72.8 4.3 81.3 3.6 66.2 2.7 44.2 5.8 25.7 2.73 62.6 2.8 66.7 2.5 49.9 4.6 38.7 2.7 21.3 1.94 81.2 3.4 86.3 3.7 76.1 5.9 48.1 7.6 31.7 5.15 54.5 2.7 56.8 3.4 42.4 1.6 28.7 3.4 16.4 1.46 56.3 1.8 61.9 5.9 49.2 6.7 36.2 5.2 19.9 3.37 58.7 3.0 67.2 1.9 64.4 2.7 35.7 4.3 24.8 1.48 67.1 6.9 84.9 4.3 71.8 4.1 51.1 3.3 29.6 2.79 59.5 3.5 66.0 3.1 52.7 6.0 35.4 3.3 20.3 2.4
Figure 4.27: Residual strength
78
European Standard (de Normalisation, 2005) provided residual strength reduction factors of
normal weight concrete at elevated temperatures, as shown in Figure 4.28. The results of resid-
ual strength reduction factors illustrate that experimental results have high residual strength
reduction factors compared to European Standard. However, the overall trend of residual
strength is the similar in a range of 1.12 to 0.33 exposed to between 600 to 800◦C.
The XRD pattern of Mix 1 is shown in Figure 4.29, where main peaks have been iden-
tified. Typical peaks are associated to Portlandite (Ca(OH)2), Calcium Silicate Hydrated
(Ca1.5·SiO3.5·xH2O) and Dicalcium Silicate(Ca2SiO4). After Mix 1 was exposed to high tem-
peratures, some of these peaks disappeared or reduced in the intensity. It is clear that in
samples exposed to 600◦C compared to 23◦C, the main hydration products such as Portlandite,
CSH disappeared. However, between 23◦C to 400◦C high temperatures, the intensity of main
peaks in XRD patterns are not clear. The reduction and dehydration of CSH can explain CSH
hydration product reform to Calcium Oxide (CaO) but less proportion (Alonso and Fernandez,
2004; Handoo et al., 2002; Xu et al., 2001). The SEM images of Mix 1 show that there is a
limited amount of fabric structures in the mixture after exposed to 600◦C compared to the
image of the mixture at 23◦C, as shown in Figure 4.30.
Figure 4.28: Residual strength reduction factors (adopted after (de Normalisation, 2005))
79
Figure 4.29: XRD pattern of Mix 1
Figure 4.30: SEM image of Mix 1 after exposed to high temperatures
80
The S/N ratio is analysed for residual strength at 200◦C, 400◦C, 600◦C and 800◦C and plotted
in Figure 4.31. It is observed that superplasticiser has the most effect on the residual strength
of blended cement after high temperature exposure. From ANOVA results, the contribution
of superplasticiser on the residual strength is determined to be 54.90%, as shown in Table
4.20. Sand to cementitious material ratio has the minor effect on residual strength with 5.79%
contribution. Fly ash is the second significant influence on the residual strength with 20.07%
contribution. The contribution of water to cementitious material ratio is determined as 19.24%.
Generally, the increase of water to cementitious material ratio decreases the residual strength.
This is well known as the release of water vapour could influence the chemical composition of
hydration products known as dehydration process (Handoo et al., 2002).
Figure 4.31: Effect of parameters after exposed to high temperatures
Table 4.20: ANOVA results on high temperatures exposure
Source DFa SSb MSc Contribution %
200◦Cresidualstrength
FA 2 566.3 283.14 23.59s/c 2 871 435.49 36.29w/c 2 808 404.01 33.66SP 2 155.1 77.54 6.46
400◦Cresidualstrength
FA 2 153.09 76.55 9.71s/c 2 916.17 458.09 58.10w/c 2 418.82 209.41 26.56SP 2 88.93 44.47 5.64
600◦Cresidualstrength
FA 2 124.79 62.39 19.06s/c 2 155.13 77.57 23.70w/c 2 287.23 143.62 43.87SP 2 87.54 43.77 13.37
800◦Cresidualstrength
FA 2 139.01 69.5 20.07s/c 2 40.08 20.04 5.79w/c 2 133.29 66.64 19.24SP 2 380.31 190.16 54.90
adegree of freedom bsum of square cmean square
81
However, it was found that reducing water to cementitious material ratio would reduce the
overall residual strength, especially after exposed to 800◦C high temperature. Between the
exposed temperature ranges 23◦C to 600◦C, reducing water to cementitious material ratio
would increase the residual strength but after exposed to 800◦C would decrease the residual
strength. Therefore,based on overall residual strength S/N ratio result of the residual strength,
the optimal mixing design of blended cement mixture is 20% of fly ash,1.5 of sand to cementitious
material ratio, 0.35 of water to cementitious material ratio, and 0.2% of superplasticiser. The
optimal mixing design can be determined with the predicted S/N ratio function as (Ross, 1996):
ηpredict = η +
f∑i=1
ηi − ηo (4.3.2)
where η = overall mean of S/N ratio, f = number of a factor, and ηi = the mean of S/N
ratio at the optimal level of each factor. The S/N ratio of optimal mix design and maximum
and minimum parameter levels are determined as presented in Table 4.21. With the predicted
S/N ratio, the residual strength is calculated by Equation (4.3.2), as shown in Figure 4.32.
According to this Figure, the residual strength of the minimum parameter levels (P2) has high
residual strength between 23◦C to 600◦C. After exposed to 800◦C, however, significant loss
in residual strength is indicated. Regression analysis was used to estimate the relationships
between residual strength and parameters. Through a linear regression analysis, the empirical
relationship obtained can be written as:
R(T ) = C0 + C1x1 + C2x2 + C3x3 + C4x4 + C5x1x2 + C6x1x3 + C7x2x3 (4.3.3)
where x1, x2, x3, x4 denote fly ash, sand to cementitious material ratio, water to cementitious
material ratio and Superplasticiser, respectively. The information about regression analysis is
presented in Table 4.23. It can be seen that corrected coefficient values of residual compressive
strength at 200◦C, 400◦C, 600◦C and 800◦C are 0.957 (95.7%), 0.955 (95.5%), 0.952 (95.2%)
and 0.952 (95.2%), respectively. These corrected coefficients indicate that the empirical model
represents the good fit of the model being validated.
Table 4.21: Predicted S/N ratio
Index FA s/c w/c SP S/N ratio
P1 Optimal 20% 1.5 0.35 0.2% 75.97P2 Min Level 0% 0 0.3 0% -93.01P3 Max Level 20% 2.5 0.4 0.2% 73.14
Table 4.22: Coefficient of empirical relationship in Equation (4.3.3)
CoefficientTemperature (T )
200◦C 400◦C 600◦C 800◦C
C0 325.51 329.07 165.98 -180.6C1 -475.71 -767.82 -153.39 270.16C2 -90.575 -85.594 -44.622 63.617C3 -706.61 -769.5 -361.94 603.37C4 4435.1 6791.2 4038.6 9894.7C5 -8.719 -15.832 -9.063 107.97C6 1420 2355.7 540.29 -1258.6C7 242.49 230.12 121.57 -217.66
82
Table 4.23: Regression analysis of residual compressive strength
(a) 200◦C
Parameters Estimate Standard error T value
Intercept 325.51 174.34 1.867FA -475.71 1203.3 -0.395s/c -90.575 56.725 -1.597w/c -706.61 565.72 -1.249SP 4435.1 13064 0.340
FA×s/c -8.719 128.89 -0.068FA×w/c 1420 3364 0.422s/c×SP 242.49 190.69 1.272
Corrected coefficient = 0.957
(b) 400◦C
Parameters Estimate Standard error T value
Intercept 329.07 174.64 1.884FA -767.82 1205.4 -0.637s/c -85.594 56.824 -1.506w/c -769.5 566.7 -1.358SP 6791.2 13087 0.519
FA×s/c -15.832 129.12 -0.123FA×w/c 2355.7 3369.8 0.699s/c×SP 230.12 191.02 1.205
Corrected coefficient = 0.955
(c) 600◦C
Parameters Estimate Standard error T value
Intercept 165.98 95.754 1.733FA -153.39 660.91 -0.232s/c -44.622 31.156 -1.432w/c -361.94 310.71 -1.165SP 4038.6 7175.2 0.563
FA×s/c -9.063 70.794 -0.128FA×w/c 540.29 1847.6 0.273s/c×SP 121.57 104.73 1.161
Corrected coefficient = 0.952
(d) 800◦C
Parameters Estimate Standard error T value
Intercept -180.6 98.353 -1.836FA 270.16 678.85 0.398s/c 63.617 32.001 1.988w/c 603.37 319.15 1.891SP 9894.7 7370 1.343
FA×s/c 170.97 72.716 1.485FA×w/c -1258.6 1897.8 -0.663s/c×SP -217.66 107.58 -2.023
Corrected coefficient = 0.952
83
Figure 4.32: Predicted residual strength
The main purpose of this section is to identify affecting parameters and optimal mix propor-
tion of blended cement for high temperature exposure. The proposed optimal mix proportion
when designed for high temperature exposure is: 20% of fly ash, 1.5 of sand to cementitious
material ratio, 0.35 of water to cementitious material ratio and 0.2% of superplasticiser. In or-
der to analyse the linear regression, the empirical relationship of residual compressive strength
was obtained. The proposed empirical models are generally higher than uniaxial compressive
strength (de Normalisation, 2005). The correct empirical cylindrical compressive strength can
be determined approximately by (de Normalisation, 2005):
fck = 0.4381 · f1.135ck,cubic + 3.123 (4.3.4)
where fck is a residual uniaxial cylindrical compressive strength, fck,cubic is a cubic compressive
strength.
4.4 Chapter Summary
This Chapter presents indentation properties of OPC and statistical analysis of properties of
blended cement. Properties of hydration products of OPC were determined using nanoinden-
tation. Deconvolution technique was used to determined indentation modulus (M), hardness
(H) and packing density (η) of hydration phases. M , H and η of LD CSH are 16.787 ± 4.804
GPa, 0.704± 0.144 GPa and 0.556± 0.01, respectively. The HD CSH has M = 30.481± 4.257
GPa, H = 1.415±0.222 GPa, and η = 0.595±0.001. Indentation fracture toughness and stress-
strain curves were generated from indentation measurement. Nanoindentation was also a useful
method for investigating the mechanical behaviour of OPC such as elastic, elastic-plastic and
plastic deformation regimes on each phase. Contact relaxation modulus and creep compliance
were also analysed. Using Rheological models with indentation test, the results of contact relax-
ation modulus indicated that stresses, under constant condition, reduce insignificantly during
the initial period within 5 second only. It can therefore be concluded that stress relaxation in
84
OPC paste can be negligible. The obtained creep compliance rate shows that MP phase is the
main phase that increases the creep compliance rate of the matrix. It means that the porosity
tends to increase creep compliance. Also, long-term creep compliance rate results show useful
metric that quantifies a unique mechanical response of creep behaviour. The strength failure
can occur in MP phase because MP phase has small strain capacity compared to another phase,
thus, high capillary porosity exhibits low compressive strength. Thus, porosity is an important
factor to be considered when designing OPC mixtures in order to optimise the compressive
strength and minimise the creep behaviour.
Based on the results of the investigation conducted on blended cement mixtures using Taguhi’s
design of experiment, Taguchi’s design of experiment and ANOVA results shows an increase in
fly ash content and water to cementitious material ratio leads to a decrease in the density of
blended cement. Superplasticiser has a minor effect on the density of blended cement mixtures.
The optimal mix design was obtained as: 20% fly ash, no content of sand to cementitious
material ratio, 0.4 of water to cementitious material ratio, and 0.1% of superplasticiser. For
compressive strength development, an increase in fly ash content and sand to cementitious
material ratio decreased the compressive strength development. The optimisation of the com-
pressive strength of blended cement mixtures was found to be 20% of fly ash content, 1.5 of
sand to cementitious material ratio, 0.35 of water to cementitious material ratio and 0.2% of su-
perplasticiser. XRD and SEM analysis confirms that the main hydration product of OPC paste
is CSH phase, which increases with curing ages. SEM image also show growth of hydration
products after curing processes. With respect to resistance to high temperature, superplasti-
ciser was identified as the most significant parameter and fly ash was the second most effect on
mixtures. Increasing in fly ash and superplasticiser was found to improve the overall residual
strength, and the optimisation of mix design was 20% of fly ash, 1.5 of sand to cementitious
material ratio, 0.35 of water to cementitious material ratio, and 0.2% of superplasticiser; XRD
and SEM analysis results also confirmed the dehydration process.
This study determined that identifying the characteristics of civil engineering materials. Taguchi’s
design of experiment approach was successfully applied as a useful tool in studying the influ-
ence of parameters in cementitious matrices. The results can be analysed using the ANOVA
technique to examine the variation in the measured properties of blended cement. Moreover,
the potential impact of indentation approach will encourage consideration of small scales exam-
ination to represent the large scale testing of civil engineering structural elements. In addition,
indentation application in civil engineering will enable formulation of composite materials such
as high-performance fibre reinforced cementitious composite.
85
Chapter 5
Properties of Alkali-Activated
Cement
5.1 Introduction
Alkali-activated cement (AAC) is a potential cementitious system for sustainable development
(Caijun and Della, 2006). Its main constituent is pozzolan, which can react with alkali activator
to form binder (Shi et al., 2011). The composition of cementitious components and alkali-
activated pozzolan cement are classified according to the type of pozzolans such as fly ash,
metakaolin, soda lime glass and natural pozzolan (Caijun and Della, 2006). A number of
researchers (Pacheco-Torgal et al., 2008; Roy, 1999; Shi et al., 2011) reported the difference
between the composition of traditional Portland cement and fundamental rock-forming minerals
of the earth crust. The common chemical compositions of Portland cement and fundamental
rocks are silica and alumina which can also be found in a number of industrial wastes.
Ground industrial wastes or by-products containing aluminosilicate when mixed with rich alkalis
could form a hydraulic binder which is a type of inorganic polymer called “geopolymer” or alkali-
activated cement (AAC). Davidovits (1994a) classified different types of geopolymer according
to Si to Al ratio in the mixtures for various industry applications. Lloyd et al. (2010) reported
that calcium content in geopolymer is important for alkali mobility that may be significant to
limit the durability of embedded steel reinforcement. A precaution should be taken as some of
the ground industrial wastes or by-products such as those containing calcium aluminate and
metakaolin geopolymer have loss of strength during long-term ageing (Provis and Van Deventer,
2009). Despite extensive research in this area, the entire polymerization process of AAC is not
totally understood. It is classified as a polymer because of its huge molecule formed by a number
of smaller groups of molecules. AAC has superior mechanical, chemical and thermal properties
compared to ordinary Portland cement (OPC) (Duxson, Provis, Lukey and Van Deventer,
2007). The main benefit of AAC is that the source material is not a carbonate-bearing material;
therefore, it does not release vast quantities of CO2 as in the case of Portland cement. Turner
and Collins (2013) reported that the carbon emission of AAC concrete is around 9% less than
comparable OPC concrete as alkali activators are high carbon footprint materials. Also, high
early age strength, high chemical durability and resistance to high temperature are beneficial
properties of AAC. Normally, cementitious materials have several phases that contribute to the
mechanical properties.
This Chapter presents the mechanical and micromechanical properties of Alkali-activated fly ash
86
cement (AAFA) paste and mortar. The experimental programme conducted has been designed
according to Taguchi’s method (Roy, 2010) an efficient approach for investigating optimum
design parameters as per the required performances. The investigated parameters are density
and compressive strength of AAFA samples containing varying proportions of constituents. In
addition to the density and compressive strength, microporomechanics of AAFA samples have
been explored using Nanoindentation and statistical analysis to determine viscoelastic proper-
ties, which are contact relaxation modulus and contract creep compliance. Further, statistical
analysis of nanoindentation based on contact stiffness measurements enables to produce inden-
tation stress-strain curves of multiphase of AAFA materials. The methods used in this Chapter
can quantify the effects of the test parameters on the mechanical and micromechanical prop-
erties of the materials, and the outcome leads to a guideline for design of AAFA mixtures to
achieve required properties.
5.2 Taguchi’s Design of Experiment
Taguchi’s design experimental approach to parameters design provides design engineers with a
systematic and efficient method for investigating optimum design parameters for performance
and cost. In this research, the following parameters are considered in the mix proportions:
- Silica fume (SF)
- Sand to cementitious material ratio (s/c)
- Liquid to solid ratio (l/s)
- Superplasticiser (SP)
Three levels of the test parameters were selected as shown in Table 5.1. The Standard L9 (34)
orthogonal array (Ross, 1996) is used as shown in Table 5.2.
Table 5.1: Variation parameters and levels for AAFA mixture
Levels SF s/c l/s SP
1 0 % 0 0.6 0 %2 2 % 0.25 0.65 2 %3 4 % 2.5 0.7 4 %
Table 5.2: Standard L9 orthogonal array
Levels SF s/c l/s SP
1 1 1 1 12 1 2 2 23 1 3 3 34 2 1 2 35 2 2 3 16 2 3 1 27 3 1 3 28 3 2 1 39 3 3 2 1
87
In this experimental work, Class F (low calcium) fly ash from Collie Power Station in Australia
was used to prepare AAFA paste and mortar samples. Sand having a specific gravity of 2.6
was used as fine aggregates, and silica fume was used as a very fine pozzolan. The chemical
composition of fly ash and silica fume is presented in Table 5.3. The Loss on Ignition (LOI) of
fly ash and silica fume were 1.5% and 3.8%, respectively. The median particle size of fly ash and
silica fume (condensed silica fume) were 45µm and 5µm, respectively. The XRD pattern of the
Class F fly ash is shown in Figure 5.1. The alkali activator was prepared by dissolving NaOH
pellet in water and left for 24 hours, and Na2SiO3 (water glass) was then added and again left for
24 hours before mixing. The chemical composition of the Na2SiO3 was: Na2O = 14.7%, SiO2 =
29.4%, and water content was 55.9% by mass. The effect of Na2SiO3 to NaOH solution by mass
on compressive strength was reported that higher Na2SiO3 to NaOH ratio would increase the
compressive strength (Hardijito and Rangan, 2005). Considering the economic cost of alkaline
liquid, Na2SiO3 to NaOH ratio of 2 was used to prepare all mixtures. A naphthalene sulphonate
superplasticiser was used to improve workability. Specimens were cast in 50mm cubic moulds
and cured in a chamber at 60◦C and 70% relative humidity for 24 hours. This temperature is
commonly used for curing alkali-activated binders (Duxson, Fernandez-Jimenez, Provis, Lukey,
Palomo and Van Deventer, 2007; Hardijito and Rangan, 2005). After that, the specimens were
placed in a 23◦C controlled room until the testing age. The compressive strength and density
of AAFA samples were tested at the age of 28 days in accordance with ASTM ASTM Standard
C109 (2011). The reported results are the average of five samples.
Table 5.3: Chemical composition (wt.%)
(a) Class F fly ash
SiO2 Al2O3 CaO Fe2O3 K3O Others
65.9 24.0 1.59 2.87 1.44 4.2
(b) Silica fume
SiO2 Na2 K2O Other
89.6 0.11 0.23 10.06
Figure 5.1: X-ray diffraction pattern of low calcium fly ash
88
5.3 Mechanical Properties
Nine mixtures of AAFA samples were prepared according to Taguchi’s design of experiment as
presented in the previous section. Details of the mixtures content and the experimental results
of the density (ρ28) and compressive strength (fck,cubic,28) at 28 days are presented in Table
5.4. Signal to Noise (S/N) ratio and analysis of variance (ANOVA) can be used to determine
the relative influence of factors and their interaction to the variation of the results. ANOVA
is a statistical test which analyses variance. It is helpful in formally testing the significance of
all main factors and their interactions by comparing the mean square against an estimate of
the experimental errors at specific confidence levels. In this study, the significant value of 0.5
was used. Graphs of S/N ratio and results of ANOVA analysis are presented in Figure 5.2 and
Table 5.5.
The obtained results clearly indicate that the density is an important parameter for reducing
the self-weight of structures. The density of AAFA at 28 days is between 1521 kg/m3 to 1717
kg/m3. The results of S/N on the density also indicate that the decreasing density is affected
by the increasing the liquid to solid ratio and superplasticiser. The ANOVA results show that
sand to cementitious material ratio is the most influencing parameter on the density of AAFA
with 79% contribution. The other influencing parameters are superplasticiser and liquid to solid
ratio with approximately 10% and 9% contributions, respectively, as shown in Figure 5.3(a). It
is observed that silica fume has the least influence on the density. The optimal mix proportion
for minimisation of density is when there is no sand with 2% of silica fume, 0.7 of liquid to solid
ratio and 4% of superplasticiser.
It is observed that the 28 days compressive strength of AAFA is in a range of 6.8 to 25.3 MPa.
Table 5.4: Mix proportion with density and compressive strength test result
Mixρ28 fck,cubic,28
(kg/m3) (MPa)
1 1570 20.302 1598 15.843 1615 12.454 1509 6.795 1586 10.406 1685 9.927 1521 8.828 1602 9.929 1717 13.25
Table 5.5: ANOVA on density and compressive strength at 28 days
Source DFa SSb MSc Contribution %
Density
SF 2 790.1 395.0 2.14s/c 2 29155 14577.5 79.14l/s 2 3270.0 1635.0 8.88SP 2 3624.5 1812.2 9.84
Compressivestrength
SF 2 133.4 66.7 55.48s/c 5.7 2.9 2.38 29.15l/s 31.2 15.6 12.99 12.99SP 70.1 35.1 29.15 2.38
adegree of freedom bsum of square cmean square
89
Figure 5.2: Effect of parameters on density and compressive strength at 28 days
90
From ANOVA results, silica fume has the most significant effect on the compressive strength
with 55% contribution as shown in Figure 5.3(b). It can be seen that there is a minimal effect
of sand to cementitious material ratio on the compressive strength of about 2% contribution.
The second most important parameter is the superplasticiser content with 29% contribution.
The optimum mix proportion for 28 days compressive strength is a normal paste, i.e., no silica
fume, no sand, no superplasticiser with liquid to solid ratio of 0.6.
The main reason to obtain regression model is to uncover causes of the outcome by studying
the relationship between variables. In this research, therefore, regression analysis was used as
a tool for investigating the relationship between variables and to predict parameters outside
of the ranges of variables. Through a linear regression analysis, the empirical relationship of
AAFA mixtures at 28 days curing ages was obtained as:
Figure 5.3: Contribution of experimental parameters on (a) density (b) compressive strength
91
ρ28 = 2004.1 + 621.43x1 + 190x2 − 721.43x3 − 8175x4
+ 182.86x2x3 − 1542.9x2x4 − 11286x3x4 (5.3.1a)
fck,cubic,28 = 72.244− 36.548x1 − 100.47x2 − 85.919x3 − 863.25x4
+ 146.97x2x3 + 237.43x2x4 + 1047.1x3x4 (5.3.1b)
where x1, x2, x3, x4 denote silica fume, sand to cementitious material ratio, liquid to solid ratio
and superplasticiser, respectively. As shown in Table 5.6, provides the results which match well
with the experimental results.
Validation of predictive model can be carried out with corrected coefficients (Analla, 1998). The
information about regression analysis is presented in Table 5.7. It can be seen that corrected
coefficient values of density at 28 days is 0.995 (99.5%). The corrected coefficient of compressive
strength at 28 days has low values compared with that of the density, as 0.855 (85.5%).
Table 5.6: Experiment and Predicted density and compressive strength at 28 days
MixExperiment Predicted Ratio
ρexp fck,cubic,exp ρpred fck,cubic,predρexp fck,cubic,exp
/ρpred /fck,cubic,pred
1 1570 20.30 1571 20.69 0.999 0.9812 1598 15.84 1588 12.70 1.006 1.2483 1615 12.45 1616 12.84 0.999 0.9694 1509 6.79 1514 8.36 0.997 0.8125 1586 10.40 1591 11.97 0.997 0.8696 1685 9.92 1690 11.49 0.997 0.8637 1521 8.82 1518 8.03 1.002 1.0988 1602 9.92 1599 9.13 1.002 1.0869 1717 13.25 1714 12.46 1.001 1.063
Average 1.000 0.999StD 0.003 0.139
92
Table 5.7: Regression analysis of test data
(a) Density at 28 days
Parameters Estimate Standard error T value
Intercept 2004.1 166.82 12.013FA 621.43 437.14 1.422s/c 190 477.98 0.399w/c -721.43 265.90 -2.713SP -8175 5974.8 -1.368
s/c×l/s 182.86 741.85 0.247s/c×SP -1542.9 1854.6 -0.832l/s×SP 11286 9273.1 1.217
Corrected coefficient = 0.995
(b) Compressive strength at 28 days
Parameters Estimate Standard error T value
Intercept 72.244 51.946 1.391FA -36.548 136.12 -0.269s/c -100.47 148.84 -0.675w/c -85.919 82.797 -1.038SP -863.05 1860.4 -0.464
s/c×l/s 146.97 231 0.636s/c×SP 237.43 577.5 0.411l/s×SP 1047.1 2887.5 0.363
Corrected coefficient = 0.855
5.4 Indentation Properties
A number of researchers (Criado et al., 2008; Fernandez-Jimenez and Palomo, 2005; Garcia-
Lodeiro et al., 2011; Xu and Van Deventer, 2000) studied the characterization of alkali-activated
materials by a variety of experimental methods including XRD, SEM, DTA and TGA. Jennings
(2000), Tennis and Jennings (2000) proposed a model for the determination of two types of
calcium silicate hydrate (C-S-H) viz., high density (HD) and low density (LD) C-S-H, at different
parts of the specimen’s geometry. The content of C-S-H is determined in term of its volume
fraction of the indentation grid. Constantinides and Ulm (2004) determined the two C-S-H
types, Portlandite (CH), and unhydrated clinker using nanoindentation method. The result
shows that decalcification of C-S-H phases is the primary source of nanometre-scale elastic
modulus degradation.
Recently, Nemecek et al. (2011) studied the reaction products of AAFA paste, which was made
from NaOH solution having the liquid to solid ratio of 0.531, and was cured at 80◦C for 12 hours.
Nanoindentation and environmental scanning electron microscope (ESEM) were used, and four
phases of reaction products were found. The four phases were identified as N-A-S-H (sodium
aluminosilicate hydrate), partly-activated slag (N-A-S-H gel intermixed with slag-like particles),
non-activated slag (porous non-activated slag-like particles), and non-activated compact glass
(solid non-activated glass spheres or their relicts). These findings also reveal that N-A-S-H
is the main reaction product which is linked to the atomic scale and nanostructure and is
independent of precursor material or the temperature curing regime. N-A-S-H phase is pure
and is related to the mechanical strength of AAFA matrix. The contents of Si ions in N-A-S-H
can be increased by the presence of Si ions in the raw materials. It has also been found that
93
the increasing condensation degree of Si ions in N-A-S-H relates directly to the mechanical
strength gain (Fernandez-Jimenez and Palomo, 2003, 2005). The partly-activated slag phase
is intermixed with the slag-like particles. The non-activated slag phase is porous and contains
non-activated slag-like particles. The non-activated compact glass phase is solid, non-activated
glass sphere.
This section presents the microporomechanics of AAFA samples using nanoindentation and
statistical analysis of nanoindentation data. A background on nanoindentation and how to
determine the indentation modulus, hardness, Poisson’s ratio, cohesion, friction coefficient and
packing density of materials have been presented in Chapter 3. Details on nanoindentation test
and statistical analysis results will be presented next.
5.4.1 Sample Preparation
For nanoindentation tests, each specimen after 28 days curing was cut using a diamond saw to
obtain 10 mm cube core part. The reaction of specimens was stopped using solvent exchange
and oven dry method (Chen et al., 2014; Zhang and Scherer, 2011). Miller et al. (2008) pointed
out that it was important to reduce the surface roughness of specimen to get an accurate nanoin-
dentation results because the specimen’s surface has a significant influence on the test results
(ISO 14577-1, 2002). Therefore, fine emery paper was used to grind all specimens to reduce
the surface roughness. After that, the surface was polished using a suspension solution ranging
from 6.0µm to 0.1mum for 45 mins and with 0.05µm for another 15 mins. Nanoindentation test
was then carried out on the four sets of 10 mm cube specimens with XP (all-purpose) testing
method (applied load: 0.5mN); each set was specified 10 ± 10 grid (100 indentations) with
20µm grid spacing. Berkovich indenter was used for the entire tests.
5.4.2 Statistical Indentation
The results of nanoindentation test using deconvolution technique in terms of the cumulative
distribution function (CDF) and probability density function (PDF) of the tested AAFA speci-
mens confirm the presence of the four phases in AAFA. This finding is similar to the previously
reported results (Nemecek et al., 2011). The typical indentation load depth (P − h) curves are
illustrated in Figure 5.4. The mean indentation depth is 240 ± 7 nm in the N-A-S-H phase,
143± 3 nm in the partly-activated slag, 82± 1 nm in the non-activated slag and 54± 4 nm in
the non-activated compact glass phases.
With the hypothesis that AAFA is porous, the quadratic minimisation problem, Equation
(3.4.8), can be solved using MATLAB (2014). The results of the quadratic minimisation proce-
dure of scaling indentation modulus and hardness exhibit the properties of the reaction prod-
ucts. A typical packing density distribution of AAFA in relation to the indentation modulus
and hardness is as shown in Figure 5.5. The results of the solid properties of reaction products
of AAFA from the quadratic minimisation are presented in Table 5.8. The ranges of stiffness,
cohesion and friction angle are 84.30 to 144.57 GPa, 0.25 to 1.46 GPa and 29.13 to 37.93 de-
gree, respectively. The properties of reaction products are to be understood in the sense of the
Drucker-Prager strength model and Coulomb material model (Ulm et al., 2007).
Analysis of S/N ratio shows the effect of the test variables on the properties of the reaction
products as shown in Figure 5.6. It can be observed that generally an increase in the value
of the parameters leads to a decrease in the stiffness of the reaction products. However, an
increase in the value of the parameters such as silica fume, sand to cementitious material ratio
and superplasticiser increases the cohesion property. It could mean that an increase in silica
fume, sand to cementitious material ratio and superplasticiser could lead to an increase in the
94
Figure 5.4: Typical indentation load-depth (P − h) curve on AAFA
Figure 5.5: Packing density relationship distribution of AAFA on Mix 1
Table 5.8: Properties of the reaction products
MixStiffness Poisson’s Cohesion Friction Friction angle(GPa) ratio (GPa) coefficient (degree)
1 144.57 0.50 1.46 0.62 31.852 87.06 0.50 0.48 0.62 31.783 84.30 0.50 0.39 0.67 33.934 84.40 0.50 0.32 0.63 32.315 103.80 0.50 0.48 0.62 31.816 87.93 0.50 0.38 0.64 32.647 84.87 0.50 0.43 0.56 29.138 84.81 0.50 0.40 0.62 31.939 94.95 0.50 0.25 0.78 37.93
95
bonding between particles to particles. Based on Mohr-Coulomb yield surface theory (Labuz
and Zang, 2015), friction angle can be used to understand Drucker-Prager yield surface. This
study reveals that relationship between stiffness and cohesion have an inverse relationship,
i.e., an increase in stiffness leads to a decrease in cohesion. ANOVA results show that the
contribution of superplasticiser, silica fume and sand to cementitious material ratio are the
most significant parameters on the stiffness, the cohesion and the friction angle, respectively,
as shown in Table 5.9 and Figure 5.7. The optimal mixing design to achieve high stiffness is
thus no silica fume, no sand content, 0.6 of liquid to solid ratio and no superplasticiser content.
The optimal mixing proportion to achieve high cohesion, based on these test results, is 4% of
silica fume, 0.5 of sand to cementitious material ratio, 0.65 of liquid to solid ratio and 4% of
superplasticiser. The optimal mixing design to achieve high friction angle is 4% of silica fume,
0.5 of sand to cementitious material ratio, 0.65 of liquid to solid ratio and no superplasticiser
content.
Table 5.9: ANOVA results on properties of reaction products
Source DFa SSb MSc Contribution %
Stiffness
SF 2 483.1 241.6 15.68s/c 2 411.8 205.9 13.37l/s 2 511.1 255.5 16.59SP 2 1674.6 837.3 54.36
Cohesion
SF 2 0.3217 0.1608 30.41s/c 2 0.2505 0.1252 23.68l/s 2 0.2625 0.1312 24.81SP 2 0.2232 0.1116 21.10
Frictionangle
SF 2 0.8509 0.4254 1.91s/c 2 23.4753 11.7376 52.62l/s 2 9.4317 4.7158 21.14SP 2 10.8536 5.4268 24.33
adegree of freedom bsum of square cmean square
96
Figure 5.6: Effect of parameters on reaction properties (a) Stiffness (b) Cohesion (c) Friction angle
97
Figure 5.7: Effect of parameters on reaction properties (a) Stiffness (b) Cohesion (c) Friction angle
98
Equation (3.5.5) was used to calculate the volume fraction of N-A-S-H, partly-activated slag,
non-activated slag, and non-activated compact glass phases of AAFA mixtures. The nanoin-
dentation deconvolution results of indentation modulus (M), hardness (H), packing density (η)
and volume fraction are obtained. As expected, N-A-S-H phase is the major reaction product
and relates directly to the mechanical strength gain (Fernandez-Jimenez and Palomo, 2003,
2005). It can be seen from the obtained results in the present case that the lowest indentation
modulus, hardness and packing density are with reference to pure N-A-S-H phase, the decon-
volution results of N-A-S-H phases vary from 4.44 to 16.78 GPa for indentation modulus, 0.11
to 0.75 GPa for indentation hardness, and 6 to 53% of volume fraction. Table 5.10 summarises
the results of the deconvolution procedure of the AAFA mixtures.
ANOVA results as shown in Table 5.11 and Figure 5.8 indicate that silica fume is the most
significant effect on the indentation modulus of N-A-S-H phase with 47% contribution, and
sand to cementitious material ratio has the minimal effect on indentation modulus of N-A-S-H
phase. The best mix proportion for the indentation modulus of N-A-S-H phase is 2% silica fume,
sand to cementitious material ratio of 0.5, liquid to solid ratio of 0.65, and 0% superplasticiser
content. For the indentation hardness of N-A-S-H, superplasticiser has the most effect, i.e.,
almost 63% contribution, while the liquid to solid ratio has the least effect, with just over 4%
contribution. The optimum mix design of indentation hardness of N-A-S-H phase is 4% silica
fume, sand to cementitious of 0.25, liquid to solid ratio of 0.6 and 4% superplasticiser content.
Based on these results, the relationship between indentation modulus and hardness of N-A-S-H
phase of AAFA is an inverse relationship. Silica fume is found to have the most adverse effect on
the volume fraction of N-A-S-H phase, with almost 76% contribution. This outcome correlates
well with the compressive strength results, i.e., the best mix for the maximum volume fraction
of N-A-S-H phase is a normal paste mixture.
99
Table 5.10: Deconvolution results
(a) Indentation Modulus (M) (Unit in GPa)
MixA B C D
µ σ µ σ µ σ µ σ
1 10.20 4.54 22.25 3.42 36.33 8.69 69.58 22.112 11.76 2.84 19.45 5.83 46.21 11.41 70.47 7.213 7.90 3.30 16.11 4.76 33.40 9.33 53.07 20.694 16.32 4.72 18.72 8.00 28.93 6.90 55.05 18.215 16.41 5.67 22.02 1.40 29.32 7.91 66.25 20.596 16.25 4.22 23.56 6.90 40.51 18.89 66.43 19.207 10.50 1.69 16.56 1.98 25.28 6.47 56.62 18.158 4.44 2.47 14.29 3.02 23.48 6.99 56.51 16.919 16.78 0.98 19.23 2.75 24.43 7.60 47.61 25.36
(b) Indentation Hardness (H) (Unit in GPa)
MixA B C D
µ σ µ σ µ σ µ σ
1 0.61 0.33 2.09 0.52 4.20 3.10 10.04 4.882 0.52 0.19 1.21 0.55 4.75 1.75 7.47 0.393 0.25 0.13 0.82 0.45 3.01 1.33 6.60 1.604 0.53 0.24 1.42 0.35 1.56 0.74 5.45 2.765 0.75 0.28 0.91 0.43 1.71 0.96 7.18 2.236 0.62 0.27 1.25 0.51 4.05 1.95 8.15 0.497 0.52 0.10 0.77 0.11 1.11 0.57 6.24 2.848 0.11 0.08 0.69 0.17 1.22 0.74 6.46 2.099 0.57 0.12 0.91 0.10 1.22 0.59 0.58 3.16
(c) Packing density (η)
MixA B C D
µ σ µ σ µ σ µ σ
1 0.55 0.02 0.61 0.01 0.61 0.12 0.72 0.052 0.58 0.02 0.63 0.03 0.77 0.05 0.83 0.013 0.56 0.02 0.61 0.03 0.71 0.05 0.80 0.044 0.60 0.03 0.64 0.02 0.69 0.02 0.82 0.075 0.60 0.03 0.62 0.01 0.66 0.04 0.82 0.056 0.61 0.03 0.63 0.04 0.75 0.07 0.86 0.017 0.58 0.01 0.62 0.01 0.65 0.04 0.84 0.078 0.53 0.02 0.61 0.02 0.64 0.05 0.83 0.059 0.60 0.03 0.62 0.01 0.63 0.04 0.76 0.08
A: N-A-S-H, B: Partly-activated slag
C: Non-activated slag, D: Non-activated compact glass
100
Table 5.11: Deconvolution results of Volume fraction (f) (Unit in %)
Mix A B C D
1 53 8 14 242 48 27 18 73 24 26 29 214 29 16 25 315 28 9 30 326 42 23 25 107 6 31 33 318 9 35 31 259 8 27 26 39
A: N-A-S-H, B: Partly-activated slag
C: Non-activated slag, D: Non-activated compact glass
Table 5.12: ANOVA results on properties of reaction products
Source DFa SSb MSc Contribution %
Modulus(GPa)
SF 2 74.06 37.1 46.95s/c 2 11.54 5.8 7.32l/s 2 34.66 17.4 21.97SP 2 37.50 18.8 23.77
Hardness(GPa)
SF 2 0.090 0.044 28.35s/c 2 0.014 0.007 4.43l/s 2 0.014 0.007 4.34SP 2 0.200 0.100 62.87
Volumefraction (%)
SF 2 0.190 0.100 75.83s/c 2 0.003 0.002 1.27l/s 2 0.036 0.020 14.26SP 2 0.022 0.010 8.62
adegree of freedom bsum of square cmean square
101
Figure 5.8: Contribution of experimental parameters on N-A-S-H phase at 28 day
102
The total activated reaction phases can also provide an estimation of the activation degree in
the form of:
ξ =
ϕ∑i=1
γfiε
(5.4.1)
where ϕ is the number of activated reaction phases, γ is the fly ash mass in the mixture, and
ε is the total mass of the mixture. Thus, the total activation degree in the present case is the
sum of the volume fraction of the reaction phases which are N-A-S-H and partly-activated slag
phase. Mercury intrusion porosimetry (MIP) is a common measurement for a valid estimation
of the pore size distribution of porous solid (Diamond, 2000). However, the measurement of
porosity with MIP appears to be valid with limited application and difficulty in estimating
the porosity of a cementitious based material. A way to determine porosity with statistical
indentation technique is by a new nonintrusive way (Ulm et al., 2007). Therefore, it is possible
to calculate the total porosity of the cementitious materials from:
φ =N∑i=1
fi (1− ηi) (5.4.2)
In the present case, the degree of activation and the porosity are found ranging from 23 to
61% and 30 to 40%, respectively which are good agreement with literature (Fernandez-Jimenez
et al., 2006). From ANOVA results as shown in Table 5.12 and Figure 5.9, silica fume has the
most significant effect on the activation degree and the porosity with 49% and 60% contribu-
tions, respectively. Superplasticiser has a minor effect on the degree of activation with 10%
contribution, and sand to cementitious material ratio has the least effect on the porosity with
less than 1% contribution. The optimum mix proportion in terms of the degree of activation of
AAFA is 4% silica fume, sand to cementitious material ratio of 0.25, liquid to solid ratio of 0.6,
and 0% superplasticiser content. For the porosity, the optimum mix is 2% silica fume, sand to
cementitious material ratio of 0.25, liquid to solid ratio of 0.7 and 4% superplasticiser content.
Table 5.13 shows a summary of degree of activation and porosity of AAFA mixtures.
Table 5.13: ANOVA results on properties of reaction products
Source DFa SSb MSc Contribution %
Degree ofactivation
SF 2 0.06697 0.033486 49.12s/c 2 0.02721 0.013603 7.32l/s 2 0.02831 0.014156 19.96SP 2 0.01384 0.006921 10.15
Hardness(GPa)
SF 2 0.004125 0.002063 60.39s/c 2 0.000055 0.000028 0.81l/s 2 0.001927 0.000964 28.22SP 2 0.000723 0.000361 10.58
adegree of freedom bsum of square cmean square
103
Figure 5.9: Contribution of experimental parameters on activated degree and porosity
Table 5.14: Degree of activation and porosity (Unit in %)
Mix Degree of activation Porosity
1 61 402 59 353 34 334 43 305 29 316 43 327 35 308 34 339 23 32
104
A refined analysis of the volume fraction is as shown in Figure 5.10 and Figure 5.11. These
Figures show the overall volume fraction of the four reaction products, N-A-S-H, partly activated
slag, Non-activated slag and Non-activated compact slag. The increase of the silica fume and
liquid to solid ratio entails a decrease in N-A-S-H phase. In turn, this decrease is due to a
decrease in the similar portion of porosity of N-A-S-H phase. While the silica fume and liquid
available for reaction decrease the amount of N-A-S-H, it also decreases the N-A-S-H porosity.
The increase in the amount of N-A-S-H phase favours the formation of loose packed N-A-S-H
phase.
Figure 5.10: Volume fraction distribution with varying silica fume of (a) reaction products (b)porosity
105
Figure 5.11: Volume fraction distribution with varying liquid to solid ratio of (a) reaction products(b) porosity
106
Of particular interest is the relationship between the compressive strength and the volume frac-
tion of the N-A-S-H phase of the AAFA mixtures, as shown in Figure 5.12. The compressive
strength increases only slightly with the increase in the volume fraction of N-A-S-H phase from
0.05 to 0.42. When the volume fraction of N-A-S-H is over 0.42, the strength increases signif-
icantly. This finding is similar to those reported by other researchers (Fernandez-Jimenez and
Palomo, 2003, 2005), i.e., the N-A-S-H phase relates directly to the mechanical strength. From
Figure 5 12, the result of the relationship indicates that in order to obtain high strength AAFA,
the volume fraction of N-A-S-H phase should be greater than 0.50. The relationship between
the compressive strength and the degree of activation at the age of 28 days is also plotted as
shown in Figure 5.13. The compressive strength increases with the degree of activation. When
the degree of activation is greater than 0.5, the compressive strength increases exponentially.
Thus, to optimise the compressive strength of AAFA mixtures, the aim should be to ensure the
degree of activation of around 0.6.
Figure 5.12: Relationship between compressive strength and volume fraction of N-A-S-H phase
Figure 5.13: Relationship between compressive strength and degree of activation
107
5.5 Viscoelastic Properties
Material has viscoelastic characteristics when its behaviour exhibits both elastic and viscous
properties. The behaviour of viscoelastic materials can be seen when, upon loading beyond the
yield point, a slow and continuous increase of strain at a decreasing rate can be observed, and
when removing stresses, strain decreases following an initial elastic recovery path showing some
energy dissipation in a form of hysteresis loop. Viscoelastic materials have a strain rate that is
dependent on time (Findley and Davis, 2013), the effect of which can be critical when designing
materials for Civil Engineering applications. Nanoindentation is one of new advanced tools to
obtain quantitative viscoelastic properties such as relaxation and creep using indentation test
results. The details of the calculated viscoelastic properties of AAFA mixtures will be presented
next.
5.5.1 Contact Relaxation Modulus
Rheological models such as Maxwell, Kelvin-Voigt and Maxwell-Kevin-Voigt models, are clas-
sical linear viscoelastic properties, which can be determined from the indentation test of AAFA
samples. In the process of classical linear viscoelastic solutions, the holding regions at the
maximum load on the obtained nanoindentation load-displacement curves were fitted to the
Equation (3.8.20) to (3.8.22) using MATLAB with nonlinear least square method (MATLAB,
2014). Table 5.14 summarises the average classic linear viscoelastic properties instantaneous
modulus Mo, spring of stiffness Mv, and unit of viscosity η of AAFA mixtures.
Table 5.15: Rheological properties (Unit in GPa)
MixMaxwell Kelvin-Voigt Maxwell-Kelvin-Voigt
Mo ηM Mo Mv ηv Mo Mv ηM ηv
1 12.27 1432.35 15.20 147.05 912.96 15.25 304.16 53994.85 665.532 11.13 1229.94 13.91 123.32 818.98 14.10 257.13 45915.57 582.043 16.61 1875.54 20.41 184.39 1280.48 20.09 445.36 5503615.74 967.734 16.35 1948.27 19.48 178.23 1273.69 19.72 433.10 14047286.98 799.325 15.72 1820.17 18.72 167.18 1089.61 18.46 347.65 453469.74 817.956 18.46 347.65 16.09 170.15 979.85 16.03 314.00 442039.30 779.667 12.53 1395.27 14.73 136.50 767.62 15.08 260.73 170675.34 556.738 11.97 1326.49 15.45 139.25 882.61 15.36 293.61 1628124.46 597.879 13.59 1488.22 17.18 168.09 944.17 18.01 357.05 2399192.36 726.76
108
The normalised contact relaxation modulus, which is defined as the relaxation modulus divided
by the instantaneous modulus, represents an important material property of linear viscoelastic
materials. The mean normalised contact relaxation modulus of Combined Maxwell-Kelvin-
Voigt is as shown in Figure 5.14. The contact relaxation modulus based on the Combined
Maxwell-Kelvin-Voigt model of contact normalised relaxation modulus presents that AAFA
mixtures have a very small initial reduction of relaxation modulus within around the first 5
second period. These results generally indicate that AAFA mixtures undergo similar normalised
relaxation modulus to OPC. In Chapter 4, the results of the relaxation modulus of OPC also
show a small reduction in the relaxation modulus, which occurs in the initial period only. Thus,
the stress relaxation and reduction of the elastic modulus of AAFA are similar to that of OPC
and are negligible during the early curing ages.
109
Figure 5.14: Normalised relaxation modulus
110
5.5.2 Contact Creep Compliance
The creep behaviour of cementitious material is largely related to the viscoelastic response
of the vital reaction products and binding phases of hardening. For long-term contact creep
compliance, the change in depth of the creep phase was fit with logarithmic function in Equation
(3.8.32) using MATLAB (2014). The average corrected coefficient and coefficients of logarithmic
curve fitting obtained are as shown in Table 5.16. The results of the curve fitting show the
corrected coefficient is over 0.98 (98%). The long-term contact creep modulus C is one of the
important parameters to determine the long-term creep compliance rate. The long-term contact
creep compliance can be determined as:
Jc (t) =1
Ctwhere C =
πPmax4 tan (α)hmaxx1
(5.5.1)
As explained in Section 3.8, Pmax, hmax and α obtained from indentation test and geometry
properties. Mix 2 and Mix 4 have the lowest and highest contact creep modulus of 336.803 and
2172.179 GPa, respectively. Figure 5.15 shows the average long-term contact creep compliance
rate of AAFA mixtures.
Based on contact creep modulus, a statistical S/N ratio analysis was performed to determine
the effect of parameters on the creep modulus of the mixtures, as shown in Figure 5.16, which
can be seen that an increase of sand to cementitious material ratio in mixtures could decrease
the creep modulus. ANOVA was performed on the results of creep modulus from nine mixtures,
as shown in Table 5.17 and Figure 5.17. ANOVA results indicate that silica fume has 26.15%
contribution, sand to cementitious material ratio 20.65%, liquid to solid ratio 22.37%, and
superplasticiser 30.83%. Therefore, the affecting parameters on the creep modulus indicate
that four parameters contribute over 20% to the creep modulus of AAFA mixtures. An increase
in sand to cementitious material ratio clearly leads to a decrease in the creep modulus. The
optimal mixing design for the creep modulus of AAFA mixtures is, thus, 2% of silica fume, no
sand content, 0.65% liquid to solid ratio and 4% of superplasticiser.
From the test results of long-term creep compliance rate, the specific creep at one year was
determined as shown in Table 5.18. The values generally agree well with literature. Wallah and
Rangan (2006) used a conventional method to determine the specific creep of AAFA concrete
at one year and reported values being in a range of 15 to 29 microstrain/MPa for 40 to 67 MPa
of compressive strength of concrete.
The specific creep results generally show that AAFA mixtures undergo lesser creep compared
to OPC. In Chapter 4, the specific creep at one year of OPC paste was observed as 18.32
Table 5.16: Average logarithmic coefficients and corrected coefficient
Mixx1 x2 x3 x4 C
R-square(nm) (s) (nm/s) (nm) (GPa)
1 11.935 3.037 -0.005 0.136 386.756 0.9862 10.395 3.123 0.015 0.151 336.803 0.9883 17.626 4.437 0.035 0.145 504.181 0.9814 7.577 14.485 -0.058 0.151 2172.179 0.9825 9.159 71.840 -0.089 0.139 536.120 0.9836 8.236 3.412 -0.074 0.151 377.457 0.9827 6.365 2.160 -0.051 0.135 397.848 0.9878 14.230 2.508 0.117 0.152 503.422 0.9879 8.057 2.497 -0.051 0.113 486.692 0.984
111
Figure 5.15: Contact creep compliance rate
112
Figure 5.16: Effect of parameters on creep modulus
Table 5.17: ANOVA results on creep modulus
Source DFa SSb MSc Contribution %
Degree ofactivation
SF 2 706711 353356 26.15s/c 2 557892 278946 20.65l/s 2 604564 302282 22.37SP 2 833046 416523 30.83
adegree of freedom bsum of square cmean square
Figure 5.17: Contribution of experimental parameters on creep modulus
113
Table 5.18: Specific creep
MixSpecific creep after one year
(microstrain/MPa)
1 19.3892 22.2653 14.8734 3.4525 13.8676 19.8677 18.8488 14.8969 15.408
microstrain/MPa for mixture having the compressive strength of 95MPa. It seems that AAFA
paste (Mix 1) has higher specific creep at one year than OPC paste, but the compressive strength
is relatively low. Thus, the specific creep of at one year of AAFA paste is low while considering
a relative compressive strength. Similarly, Warner et al. (1998) presented that specific creep
of OPC concrete after one year was around 50 to 60 microstrain/MPa for 60MPa concrete,
30 to 40 microstrain/MPa for 80 MPa concrete, and 20 to 30 microstrain/MPa for 90 MPa
concrete. That means, generally, the specific creep of AAFA is less than the values reported in
the literature. According to Wallah and Rangan (2006), the reason for smaller creep behaviour
of AAFA compared to OPC may be due to “block-polymerisation”, that is, the silicon and
aluminium composition in the fly ash are not completely dissolved in the alkaline solution. The
polymerisation takes place only on the surface of the atoms and is adequate to form the blocks
to produce the polymer binder.
As presented earlier in Section 5.4.2, the deconvolution technique identified material phases
as N-A-S-H, Partly-activated slag, Non-activated slag, and Non-activated compact glass. In
addition, this deconvolution technique can identify creep modulus C for each phase. Table
5.19 shows results of M , H, C and η with four Gaussian deconvolution input parameters. The
deconvolution results of creep modulus show that N-A-S-H is mainly the phase that increases the
creep compliance of AAFA. Due to “block-polymerisation”, Partly-activated and Non-activated
phases in which the silicon and aluminium composition are not completely dissolved in the
alkaline solution show having smaller creep behaviour because of higher creep modulus than
that of the N-A-S-H phase. Therefore, Partly-activated and Non-activated phases are leading
the creep behaviour of AAFA due to high creep modulus.
The S/N ratio was analysed to understand the effect of parameters on the creep modulus of
Partly-activated and Non-activated phases. As shown in Figure 5.18, liquid to solid ratio is the
most significant parameter on creep modulus. An increase in liquid to solid ratio leads to a
decrease in the creep modulus. Therefore, an increase in liquid to solid ratio could be a cause
of increasing creep behaviour of AAFA. ANOVA was also performed on the results of creep
modulus of Partly-activated and Non-activated phases as shown in Table 5.20 and Figure 5.19.
The results show that contribution of liquid to solid ratio on creep modulus of Partly-activated
slag, Non-activated slag and Non-activated compact glass are 50.74%, 88.69% and 68.08%,
respectively. Also, it was found that sand to cementitious material ratio and superplasticiser
have a minor effect on creep behaviour of AAFA.
114
Table 5.19: Deconvolution results of Creep Modulus (C) (Unit in GPa)
MixA B C D
µ σ µ σ µ σ µ σ
1 61.597 38.354 200.611 78.705 1185.385 742.521 11946.131 5933.3132 78.028 35.566 177.069 80.719 658.149 403.671 5451.601 5933.3133 40.982 29.609 120.580 60.085 526.086 270.479 2757.519 1485.6644 63.828 28.705 212.993 89.196 853.733 508.367 6518.359 4303.5785 65.646 35.395 206.029 61.857 737.200 396.137 10523.844 6704.9336 98.314 60.987 296.108 119.977 1415.070 704.133 12209.248 6868.1917 92.288 37.273 195.745 53.218 628.075 303.435 4371.304 3105.2108 105.965 67.850 258.409 169.916 1488.145 830.584 11545.101 6116.9269 65.515 23.356 180.902 60.882 519.336 349.451 5281.275 4111.733
A: N-A-S-H, B: Partly-activated slag
C: Non-activated slag, D: Non-activated compact glass
Table 5.20: ANOVA results on creep modulus of partly-activated and non-activated phases
Source DFa SSb MSc Contribution %
Partly-activatedslag
SF 2 8017.5 4008.8 40.47s/c 2 344.6 172.3 1.74l/s 2 10051.4 5025.7 50.74SP 2 1395.8 697.9 7.05
Non-activatedslag
SF 2 68104 34052 5.99s/c 2 29827 14913 2.62l/s 2 1008916 504458 88.69SP 2 30730 15365 2.70
Non-activatedcompactglass
SF 2 16521193 8260597 15.19s/c 2 9059186 4529593 8.33l/s 2 74028816 37014408 68.08SP 2 9133741 4566871 8.40
adegree of freedom bsum of square cmean square
115
Figure 5.18: Effect of parameters on creep modulus of partly-activated and non-activated phases
116
Figure 5.19: Contribution of experimental parameters on creep modulus of (a) Partly-activated slag(b) Non-activated slag (c) Non-activated compact glass
117
5.5.3 Indentation Stress-Strain Curve and Fracture Toughness
The geometry of the indenter tip and the indented material are relating to the maximum
indenting depth h and contact depth hc. The function, hc = f (h), can be determined by contact
stiffness measurement (CSM) between the indented material and the indenter process using
Equation (3.2.13). Knowing hc = f (h), and the geometry of indenter tip, the instant contact
area can be determined by the instant stress from the loading process using Equation (3.9.1).
According to Equations (3.9.3) to (3.9.5), the indentation stress-strain curve can be obtained
from statistical deconvolution phases of AAFA paste (Mix 1) as presented in Section 5.4.2. The
indentation stress-strain curves of statistical deconvolution phases present that each phase can
be captured based on initial and stationary yield using Berkovich indenter tip with 20nm of the
radius. The results of indentation stress-strain curves clearly illustrate elastic, elastic-plastic
and plastic regimes. The initial yielding points are observed approximately when the modulus
values are 103 MPa, 87 MPa, 240 MPa and 987 MPa for N-A-S-H, Partly-activated slag, Non-
activated slag and Non-activated compact glass phases, respectively. The stationary yielding
points from indentation stress-strain curve are obtained when the strength values reach 412
MPa, 386 MPa, 1492 MPa and 1975 MPa for N-A-S-H, Partly-activated slag, Non-activated slag
and Non-activated compact glass phases, respectively. Figure 5.20(a) presents the indentation
stress-strain curve of N-A-S-H phase which shows that it remains stable in the fully plastic
regime. The indentation stress-strain curves of Partly-activated slag, Figure 5.20(b), Non-
activated slag, Figure 5.20(c), and Non-activated compact glass phase, Figure 5 20(d), illustrate
that indentation strain hardening increases in the fully plastic regime and the fracture point
cannot be observed.
Figure 5.20: Indentation stress-strain curve
118
The fracture toughness of a material is related to the energy stored in the form of elastic or
plastic energy before fracturing. A robust method for determining fracture toughness with
nanoindentation was introduced in Chapter 3.6, using Equations (3.6.8) to (3.6.8). The overall
indentation results of the fracture energy release rate (Gc) and the fracture toughness (Kc)
were determined. The results are as shown in Table 5.21.
Based on the fracture toughness results, a statistical S/N ratio analysis was performed to
determine the effect of parameters on fracture toughness of AAFA mixtures, as shown in Figure
5.21, which can be seen that increase of silica fume in mixtures could increase fracture toughness.
ANOVA was performed on the results of the creep modulus from the nine mixtures as shown
in Table 5.22 and Figure 5.22. ANOVA results show that silica fume has 29.60% contribution,
sand to cementitious material ratio 20.41%, liquid to solid ratio 22.28%, and superplasticiser
26.71%. Therefore, the affecting parameters on the fracture toughness designate that the four
parameters contribute over 20% to fracture toughness. The increase of silica fume clearly shows
an increase in fracture toughness. The optimal mixing design for fracture toughness is, thus,
4% of silica fume, 0.25 of sand to cementitious material ratio, 0.65% liquid to solid ratio and
2% of superplasticiser.
Table 5.21: Fracture energy release rate and toughness
MixGc Kc
(N/m2) (MPa m1/2)
1 0.960 0.5352 0.709 0.4003 1.176 0.6284 0.817 0.4815 0.744 0.4746 0.731 0.4467 0.752 0.4428 0.857 0.4249 0.659 0.438
Figure 5.21: Effect of parameters on fracture toughness
119
Table 5.22: ANOVA results on fracture toughness
Source DFa SSb MSc Contribution %
Fracturetoughness
SF 2 0.011415 0.005707 29.60s/c 2 0.008257 0.004128 21.41l/s 2 0.008594 0.004297 22.28SP 2 0.010300 0.005150 26.71
adegree of freedom bsum of square cmean square
Figure 5.22: Contribution of experimental parameters on fracture toughness
5.5.4 Chapter Summary
The results obtained from this investigation are based on Taguchi’s experimental design ap-
proach and statistical analysis of nanoindentation results of AAFA. Taguchi’s design of exper-
iment with ANOVA technique in a cement-based binder was successfully applied. The results
were then used to determine the effect of the test parameters on the compressive strength, den-
sity, and indentation properties such as deconvolution phases, viscoelastic, stress-strain, and
fracture toughness of AAFA. Based on the results of this study, the following conclusions can
be drawn:
- In terms of compressive strength, the normal AAFA paste, i.e., no silica fume, no sand,
no superplasticiser with liquid to solid ratio of 0.6 is the optimum mix. For the four
parameters investigated, silica fume has the most adverse impact on the compressive
strength. Increases of superplasticiser and liquid to solid ratio also contribute to the
decrease in compressive strength. There is no significant effect of sand to cementitious
material ratio on the compressive strength. In terms of density, the increase in the sand
content significantly increases the density of the mixture. The increasing of liquid to solid
ratio and superplasticiser dosage further decreases the density of AAFA mixture.
- An application of nanoindentation technology was applied for the micro poromechanics of
AAFA mixtures. Statistical analysis based on deconvolution technique was carried out to
determine the mechanical properties. AAFA was considered as the porous material. Thus,
the packing density of each phase could be estimated by solving a quadratic minimisation.
Properties of reaction products were determined based on Drucker-Prager strength and
Coulomb material models. Stiffness and cohesion of reaction products have an inverse
120
relationship.
- Four main phases of the reaction products viz., N-A-S-H, Partly-activated slag, Non-
activated slag and Non-activated compact glass phases were identified. Other properties of
the reaction products such as stiffness, Poisson’s ratio, and cohesion and friction coefficient
were also determined.
- The analysis confirms that N-A-S-H phase is the major reaction product of AAFA. The
compressive strength is shown to relate to the volume fraction of N-A-S-H phase. To
obtain high strength AAFA, the volume fraction of N-A-S-H phase should be larger than
0.42. The volume fraction of the N-A-S-H phase of 0.50 or over is recommended.
- The compressive strength is also shown to relate to the activated degree of AAFA. When
the degree of activation is greater than 0.5, the compressive strength increases exponen-
tially. Thus, it is recommended that the degree of activation should be around 0.6 to
obtain high strength mixture.
- The relaxation modulus of AAFA was found to have similar behaviour of that of OPC,
i.e., only small initial reduction occurred within a very short period (5 second). Thus,
the stress relaxation and the reduction of the elastic modulus of AAFA can be neglected
during early curing ages.
- The creep behaviour study revealed that Partly-activated and Non-activated phases are
the main reason for creep in AAFA due to “block-polymerisation” concept. It was also
found that liquid to solid ratio is the most affecting parameter on creep, an increase
of liquid to solid ratio leads to more creep. Sand to cementitious material ratio and
superplasticiser have a minor effect on creep behaviour.
- A robust method was used to determine fracture toughness. It was obtained that an
increase in silica fume leads to an increase in fracture toughness. Overall, the fracture
energy release rate and the fracture toughness of AAFA are 0.822 N/m2 and 0.474 MPa
m1/2, respectively.
In general, the study presented in this Chapter shows that mechanical and micromechanical
properties of AAFA can be obtained by means of statistical analysis of nanoindentation test
data.
121
Chapter 6
Strain-hardening Behaviour of
Cementitious Composite
6.1 Introduction
Previous Chapters presented cementitious materials properties such as mechanical, chemical,
microstructure and nano-scale properties. This Chapter presents a study on properties of High-
Performance Fibre Reinforce cementitious composite (HPFRCC).
Cementitious materials, such as mortar and concrete, generally show brittle tensile behaviour.
However, the brittle tensile behaviour could be significantly improved by adding discontinuous
fibres. Historically, general reinforcement in concrete has been in the form of continuous rein-
forcing bars, which should be in an appropriate location to resist the imposed tensile and shear
stresses. In fibre reinforced cementitious composite, fibres are discontinuous and are randomly
distributed throughout the cementitious matrix. They tend to be more closely located than
conventional reinforcing bars and are therefore better at controlling cracking. High-Performance
Fibre Reinforced Cementitious Composite (HPFRCC) is a type of material that exhibits pseudo
strain-hardening characteristic under uniaxial tensile stress among short discontinuous fibre re-
inforced cementitious composites. The “High-Performance” is the quality in fibre reinforced
cementitious composite based on the shape of its stress-strain curve in direct tension (Naa-
man and Reinhardt, 2008). Figure 6.1 illustrates strain softening and pseudo-strain hardening
(stress-strain hardening) responding behaviour after first cracking occurs.
HPFRCC can be generally classified by composite mechanics, energy and numerical approach.
One way to define the condition to accomplish strain hardening behaviour is that post-cracking
strength of the composites is higher than its cracking strength. It is, therefore, necessary
to understand some important parameters which are related to the shape of the stress-strain
relationship of HPFRCC (Li, 1997).
Generally, the stress-strain curve of HPFRCC is as shown in Figure 6.2, the first cracking
marked as A is defined as the first visible cracking or deviation from linearity as detected along
to an initial ascending portion of the stress-strain curve. During this time, several micro-cracks
might be developed in the member or has “percolated” through the structural tensile member
(Naaman and Reinhardt, 2008). The first cracking points are termed σI
and εI, respectively.
According to Li and Leung (1992), using energy approach and fracture mechanics to model
HPFRCC, stress at flaw tip of the cracking point is relative to the stress intensity factor of
KI , which could be derived from basic principle of elasticity by the principle of linear elastic
122
Figure 6.1: The concept of stress-strain hardening and strain softening under tensile stress
fracture mechanics (LEFM). The first cracking state is assumed to be equal to the steady-
state cracking stress. Two possibilities exist for fibre reinforced composite after first transverse
crack, which are strain hardening characterised by multiple cracking, or strain softening and
localisation characterised by the continuous opening of major crack during fibre pull-out. The
peak point at the end of the strain hardening branch marked as B on Figure 6.2 is at the
maximum post-cracking stress and strain coordinates in term of σII
and εII
, respectively, For
strain hardening composite, σII> σ
I, while for strain softening composite, σ
II< σ
I. After
peak point B, no more cracks could develop and one crack becomes critical defining that the
onset of crack localisation will open under increased deformation.
Figure 6.2: HPFRCC stress-strain behaviour
123
6.2 Design of Strain-hardening Behaviour
Naaman (1972); Naaman and Reinhardt (2008); Naaman et al. (1974) developed an analytical
model of the fibre reinforced composite. The number of fibre per unit volume and unit area of
the composite are important for deriving the tensile strength of composite at first cracking. It is
assumed that fibre diameter is df . The average number of fibres per unit volume of composite,
Nv is given as:
Nv =4Vfπd2fLf
(6.2.1)
where Vf is the fibre volume fraction and Lf is the length of the fibre. The number of a unit
area, Ns, is:
Ns =4Vfπd2fLf
α2 (6.2.2)
where α2 is one for unidirectional fibres, 2/π for fibres randomly oriented in planes, and 0.5 for
randomly oriented in space. If fibres are distributed randomly, the stress cracking of the matrix
in tensile prime can be illustrated by:
σI = σmu (1− Vf ) + ατVfLfdf
(6.2.3)
σII
= λτVfLfdf
(6.2.4)
where σmu is the tensile strength of the matrix, τ is the average bond strength between the
matrix and the fibre. The coefficients α and λ are to account for the fibre distribution, orien-
tation and bond efficiency. Developing a strain-hardening behaviour, the following condition
must be satisfied:
σII≥ σ
I(6.2.5)
Substituting Equations (6.2.3) and (6.2.4) into Equation (6.2.5) (Naaman, 1972; Naaman and
Reinhardt, 2008; Naaman et al., 1974):
λτVfLfdf≥ σmu (1− Vf ) + ατVf
Lfdf
(6.2.6)
Solving Equation (6.2.6) for Vf then (Naaman, 1972; Naaman and Reinhardt, 2008; Naaman
et al., 1974):
V criticalf ≡ Vf ≥1
1 + τ/σmuLf/df (1− Vf )
(6.2.7)
This equation shows the critical volume fraction of the fibre required to achieve the strain-
hardening condition of the fibre reinforced composites.
Similarly, Li and Leung (1992) studied the condition for steady-state and multiple cracking for
a 3-D randomly distributed discontinuous fibre reinforced composite. Based on this study, Li
et al. (1995) shows that the condition to ensure strain-hardening behaviour of composite is:
V criticalf ≡ Vf ≥12Gc
gτ (Lf/df ) δo(6.2.8)
where Gc is the composite critical energy release rate. The g term represents snubbing factor.
The crack opening at fibre bridging stress reaches a maximum, δo, when
δo =τLf
Efdf (1 + η)(6.2.9)
124
where η = (VfEf ) / (VmEm), fibre volume fraction and elastic modulus are Vf , Ef , respectively,
and Vm, Em are the matrix volume fraction and elastic modulus, respectively. The snubbing
factor g is defined in term of snubbing coefficient f as:
g ≡ 2
4 + f2
[1 + exp
(πf
2
)](6.2.10)
where
f =1
Φln(P)
(6.2.11)
The term Φ is inclining angle between fibre and matrix, P is normalised force of fibre. Li et al.
(1990) studied the snubbing friction with fibre pull-out test from cementitious matrix. They
found that average nylon fibre of snubbing coefficient f is 0.994 and 0.702 for polypropylene
fibre. With fibre pull-out test, snubbing friction increases the pull-out resistance and contributes
to the overall composite toughness.
Also, the ultimate strength of the composite σcu corresponds with the maximum bridging stress
σo when a brittle matrix presents strain-hardening behaviour as (Li et al., 1990):
σcu =1
2gτVf
(Lfdf
)(6.2.12)
In Equation (6.2.8), the composite critical energy release rate is given by (Budiansky and Cui,
1994; Li et al., 1995):
Gc = (1− Vf )(1− v2m
) K2m
Em≈ K2
m
Em(6.2.13)
where vm is the matrix Poisson’s ratio, and Km is matrix fracture toughness. The strain
hardening of fibre volume fraction is generally limited to few percentages. Thus, Vf term
in equations can be eliminated. Therefore, an approximate equation for critical fibre volume
fraction required to achieve strain-hardening behaviour is:
V criticalf ≈ 12Gcgτ2
d2fL3f
(6.2.14)
Kanda et al. (2000) studied the theoretical prediction of the tensile stress-strain curve of strain
hardening composite. According to their study, the relationship between tensile stress and
strain is characterised by two stages that are first crack state and the ultimate state, as shown
in Figure 6.3. It shows that tensile stress-strain relationships can be represented by the bilinear
behaviour. The first crack state is assumed to be equal to the steady-state cracking stress σfc,
which is the initially bend-over point and initiation of multiple cracking. The ultimate state is
the peak tensile stress state as multiple cracking terminates. Thus, a bilinear presentation of
HPFRCC can be illustrated as (Kanda et al., 2000):
σ (ε) =
Ecε when ε ≤ σfc/Ec
σi + Eiε when ε > σfc/Ec
(6.2.15)
where Eie and σi are:
Eie =σpeak − σfcεcu − σfc/Ec
, σi =
(1− Eie
Ec
)where σpeak and εcu are ultimate theoretical stress and respectively, and the composite elastic
modulus Ec. The steady-state cracking stress can be expressed with composite Poisson’s ratio
125
Figure 6.3: Tensile stress-strain curve for HPFRCC (adopted after (Kanda et al., 2000))
vc by (Kanda et al., 2000):
σfc = σog[√
2cs −cs2
](6.2.16)
where
σo =Vfτ
2
Lfdf
δ∗ =2τ
Ef (1 + η)Lf
df
Ktip =KmEcEm
Ec = EmVm + EfVf
K =Ktip
gσocoδ∗co =
(LfEc2Ktip
)2π
16 (1− v2c )2 cs =
√cs
ˆdelta∗ c =
csco
And the term of cs can be obtained by solving:
K =2cs√π
(√2cs3− cs
4
)(6.2.17)
The theoretical ultimate stress σpeak can be obtained as (Kanda et al., 2000):
σpeak =Vfgτ
2
Lfdf
(6.2.18)
The ultimate theoretical strain in terms of micromechanical parameters can be written as:
εcu =δpeak
xtheoryd
(6.2.19)
where σpeak is ultimate crack opening displacement (COD) and xtheoryd is theoretical unlimited
crack spacing that is the distance normal to the crack surface for the load in the matrix.
The σpeak was predicted considering the slip-hardening occurrence in the behaviour of a single
polymeric fibre pull-out of a cementitious matrix as (Kanda et al., 2000):
δpeak =3dfβ1 − Lfβ2 + Lfβ2
[(L2
fβ22−14Lfdfβ1β2+9d2fβ
21+32dfβ1
L2fβ
22
]1/24Lfβ2
(6.2.20)
where β1 and β2 are the first and second order non-dimensional hardening parameters from the
load-displacement relation observed in the single fibre pull-out test, respectively.
The xtheoryd can be predicted in terms of saturated crack spacing xd which depends on trans-
ferred stress via bridging fibre at cracking plane to non-cracked matrix plane and flaw size
126
distribution F (cc) at cracking stress reached maximum bridging stress. The theoretical ulti-
mate crack spacing can be expressed by (Kanda et al., 2000):
xtheoryd =xd
1− F (cc)(6.2.21)
The saturated crack spacing xd can be expressed by (Kanda et al., 2000):
xd =Lf2−
[L2f −
2VmdfLfσmu
gτVf
]1/22
(6.2.22)
where σmu is tensile strength of matrix. The flaw size distribution F (cc) can be given by
Weibull function as (Kanda et al., 2000):
F (cc) = exp
[− 1
λ
(cocc
)m](6.2.23)
where reference data fitting can determine scale factor λ, m is Weibull modulus that is typically
assumed to fall between 2 to 3 for concrete. The normalised flaw size at maximum bridging
stress cc can be obtained by solving:[2√
2cc3− cc
4
]+
√π
2
K
cc= 1 (6.2.24)
Moreover, reference crack radius co = cm which is:
cm =
√cm
δ∗cm =
cmco
cm =
[√π
2
Km
σmu
]Equations (6.2.15) to (6.2.24) present the tensile stress-strain prediction. This approach can be
validated with the experimental results. Based on the theoretical discussion, Li et al. (1995)
recommended the following guideline for desirable matrix properties.
- The lower the composite critical energy release rate Gc is to achieve ductile behaviour of
the composite easily.
- The fine aggregate contents achieve upper limit for the composite critical energy release
rate.
- A higher bond strength enhances composite performance in term of composite strength.
The minimum fibre-matrix interfacial bond strength should have 0.5 MPa.
- A lower first crack strength indicates a low fibre volume fraction required. The tensile
strength of the matrix is limited to 3.0 MPa for practical purpose.
In this study, Equations (6.2.8) to (6.2.13), proposed by several authors (Budiansky and Cui,
1994; Kanda et al., 2000; Li et al., 1995, 1990), were used to design mix proportion of HPFRCC.
6.3 Single Fibre Pull-out
It is important to note that the interface properties between the fibre and matrix influence
significantly the performance of composites. The interface properties are also important in
the fracture mechanism and the fracture toughness of composite. The failure process in a
composite material when a crack propagates is complex and involves matrix cracking. The
bonding strength between fibre and matrix is to be considered as a source of energy dissipation
of HPFRCC. The single fibre pull-out test is the most common method to understand the
127
interfacial strength. Generally, fibre pull-out test has three stages during debonding (Chen
et al., 2009a,b; Herrera-Franco and Drzal, 1992; Hsueh, 1990; Kullaa, 1996; Naaman et al.,
1991; Zhan and Meschke, 2014), as shown in Figure 6.4. Each stage of single fibre pull-out test
can be expressed by:
- The first stage, So: the fibre and matrix is bonded until reach the maximum interfacial
bond strength τmax,
- The second stage, So − S1: a crack propagation could occur along the interface between
the fibre and matrix which is leading to a complete debonding,
- The third stage, S1 − Sref : fibre is pulled out from the matrix and starts to slip.
Thus, the maximum pull-out force is the most important parameter of HPFRCC, which can
present maximum interfacial bond strength. The purpose of the study in this Section is to
investigate a single PVA fibre pull-out test by FE analysis using cohesive zone materials (CZM)
model.
6.3.1 Numerical Model and Validation
A numerical study was carried out using commercial finite element (FE) software package AN-
SYS (ANSYS R○, 2015). A 2-D axisymmetric model was employed for simulation of single fibre
pull-out process. In the developed model, PVA fibre with a radius Rf was embedded at the
centre of the cylindrical matrix and Ld was the total embedded length of the fibre. The bottom
of the model was constrained in both radial and axial directions. The interface properties were
used the bilinear CZM model (MODE II), which is established by fracture mechanic models such
as interface traction and the separations, as shown in Figure 6.5. The relationship between nor-
mal critical energy Gcn and tangential critical energy Gct can be expressed by maximum normal
contact stress σmax, maximum tangential contact stress τmax, complete normal displacement
δn, and complete tangential displacement δt as (ANSYS R○, 2012):
Gcn =1
2σmaxδn (6.3.1a)
Gct =1
2τmaxδt (6.3.1b)
Figure 6.6 presents the model of the FE single fibre pull-out test. The fibre and matrix
model were meshed with 122406 six node quadrilaterals elements, as shown in Figure 6.7.
Figure 6.4: Idealised interface law in three stages for single fibre pull-out (adopted after (Zhan andMeschke, 2014))
128
Figure 6.5: Fracture Models
This model was analysed with a non-linear geometry method and convergent displacement
control as illustrated in Figure 6.8. To confirm the validity of FE analysis of single fibre
pull-out, the analytical fibre pull-out test was conducted. An interfacial friction law for the slip
Figure 6.6: Single fibre pull-out simulation model without inclined angle
Figure 6.7: Meshing configuration
129
Figure 6.8: Equivalent stress
mechanism between the fibre and the matrix has been investigated by several authors (Kullaa,
1996; Naaman et al., 1991; Zhan and Meschke, 2014). Zhan and Meschke (2014) proposed the
model based on the interfacial law that model can capture the major mechanism involved in
various situations as follows:
τ (s) =
G · S ; S ≤ So bondend stage
τmax ; So < S ≤ S1 debonding stage
τo + (τmax − τo) exp [(S1 − S) /Sref ] ; S > S1 silding stage
(6.3.2)
where
τ : interfacial shear stress
τmax : maximum interfacial bonding stress
τo : asymptotic value of frictional stress
s : relative displacement which is a point
on the fibre axis with respect to the boundary of the matrix
G : relative modulus = E/ [df (1 + vm) ln ξ]
ξ : Rm/Rf
Rm : radius of matrix
Rf : radius of fibre
Sref : parameter controlling descending branch of curve
Based on this interface law, the values of the pull-out force F and displacement δ are obtained
at different stages as:
Fo =πdfτmax
λtanh (λLd) , δo = so =
τmaxG
; for bonded (6.3.3a)
Fmax = πdfτmaxLd, δmax = So +pidfτmaxL
2d
2AfEf; for debonding (6.3.3b)
F (δ) = πdf
[τo + (τmax − τo) exp
(δmax − δSref
)](Ld − δ + δmax) ; for sliding (6.3.3c)
130
where
λ =
√πdfG
AfEf
Fe =πdfτmax
λtanh [λ (Ld − Lr)]
Lr = remaining bonded section length
Af = cross setion area of fibre
For analytical fibre pull-out, Equations (6.3.2) and (6.3.3) were used to obtain the fibre pull-
out force. The results of analytical and FE model of single fibre pull-out are overall in good
agreement, as illustrated in Figure 6.9. Thus, FE simulation can be used for investigating
interface behaviour between fibre and matrix.
Figure 6.9: Validation of FE model with analytical model
131
6.3.2 Taguchi’s Design of Experimental
Taguchi’s design experimental approach, eight parameters and three levels of test variable were
selected according to literatures (Budiansky and Cui, 1994; Kanda et al., 2000; Kullaa, 1996;
Li et al., 1995, 1990; Naaman et al., 1991; Zhan and Meschke, 2014), as shown in Table 6.1.
The Standard L27 (313) orthogonal array (Ross, 1996) is used according to these parameters.
The detail of L27 orthogonal array is shown in Table 6.2. In this numerical study, the non-
linear geometry method of displacement control for convergent was used. The displacement
was applied as the value of δt.
Table 6.1: Variation parameters and levels
Parameter Level 1 Level 2 Level 3
Elastic modulus of matrix, Em (GPa) 20 25 30Diameter of matrix, dm (mm) 5 10 15Poisson’s ratio of matrix, vm 0.2 0.22 0.25Elastic modulus of fibre, Ef (GPa) 40 120 210Diameter of fibre, df (mm) 0.038 0.5 1Fibre embedded length, Ld (mm) 4 10 12Maximum tangential traction, τmaxt (MPa) 0.5 1 1.5complete tangential displacement δt (mm) 0.1 0.25 0.4
Table 6.2: Standard L27 orthogonal array
No. Em dm vm Ef df Ld τmaxt Gct
1 1 1 1 1 1 1 1 12 1 1 1 1 2 2 2 23 1 1 1 1 3 3 3 34 1 2 2 2 1 1 1 25 1 2 2 2 2 2 2 36 1 2 2 2 3 3 3 17 1 3 3 3 1 1 1 38 1 3 3 3 2 2 2 19 1 3 3 3 3 3 3 210 2 1 2 3 1 2 3 111 2 1 2 3 2 3 1 212 2 1 2 3 3 1 2 313 2 2 3 1 1 2 3 214 2 2 3 1 2 3 1 315 2 2 3 1 3 1 2 116 2 3 1 2 1 2 3 317 2 3 1 2 2 3 1 118 2 3 1 2 3 1 2 219 3 1 3 2 1 3 2 120 3 1 3 2 2 1 3 221 3 1 3 2 3 2 1 322 3 2 1 3 1 3 2 223 3 2 1 3 2 1 3 324 3 2 1 3 3 2 1 125 3 3 2 1 1 3 2 326 3 3 2 1 2 1 3 127 3 3 2 1 3 2 1 2
132
6.3.3 Effect of Parameters on Maximum Pull-out Force
Based on the numerical study with Taguchi’s design experimental approach, a statistical S/N
ratio analysis was performed to determine the effect of these parameters on Pmax as shown in
Table 6.3 and Figure 6.10. The S/N ratio shows that diameter of fibre has the most effect on
the fibre pull-out force. The elastic modulus of fibre and matrix has a minor effect on the pull-
out force. ANOVA was conducted and its results indicate that the contribution of diameter
of fibre on pull-out force contributed is 44.69% of in Table 6.4 and Figure 6.11 present the
summary of the results. It can be observed that increasing the elastic modulus of the matrix,
the diameter of the fibre, tangential traction and embedded length of fibre result in increasing
the pull-out force. The contribution of the elastic modulus of the matrix, tangential traction
and embedded length of fibre on the pull-out force are 14.48%, 8.92% and 9.47%, respectively.
Increasing the diameter of the matrix and Poisson’s ratio resulted in decreasing in the pull-out
force. It is observed that the diameter of the matrix, Poisson’s ratio, the elastic modulus of fibre
and complete tangential displacement are minor contributing parameters on the pull-out force,
approximately 2.5% of contribution. Through regression analysis, the empirical relationship
Table 6.3: Numerical studies of single fibre pull-out with Taguchi’s DOE
No. Pmax (N) No. Pmax (N) No. Pmax (N)
1 0.23 10 1.266 19 1.4262 15.508 11 9.421 20 9.4143 56.091 12 15.705 21 12.5654 0.239 13 1.09 22 1.6995 15.666 14 9.408 23 9.4146 55.945 15 15.585 24 12.5637 0.239 16 1.356 25 1.518 15.611 17 9.404 26 9.1929 14.614 18 15.692 27 12.556
Figure 6.10: S/N ratio of single fibre pull-out
133
obtained can be written as:
Pmax = −51.1871Em + 19.9782Ef − 3.0446Emdm − 0.2965EmEf − 9.3565Emτmax
− 4.8925EmLd + 4.2477Emδt + 0.1150dmEf − 9.7633dmdf − 8.5607dmτmax
+ 0.1457dmLd + 0.65233Efδt − 69.0918vmEf + 431.227vmLd − 0.1679Efdf (6.3.4)
− 0.3646Efτmax + 0.0465EfLd + 3.0134E2m + 3.0893d2m + 0.0047E2
f + 1.9292L2d
The correct coefficient (R-square) is observed as 0.999.
Table 6.4: ANOVA of fibre pull-out force
Source DFa SSb MSc Contribution %
Pmax
Em 2 737.6 368.8 14.48dm 2 127.2 63.61 2.50vm 2 129.3 64.65 2.54Ef 2 124 61.98 2.43df 2 2276.5 1138.27 44.69τmax 2 454.4 227.19 8.92Ld 2 482.5 241.24 9.47δt 2 126.7 63.36 2.49
Error 10 653.8 63.58 12.48
adegree of freedom bsum of square cmean square
Figure 6.11: Parameter contribution on fibre pull-out test
134
6.3.4 Single Fibre Pull-out Test with Polyvinyl Alcohol (PVA) Fibre
Polyvinyl alcohol (PVA) fibres have been used commonly on a large scale in the construction
as they have good mechanical and chemical characteristics. The monofilament PVA fibre was
used in this research, its diameter and length are 38µm (8 deniers) and 8mm, respectively.
PVA fibre has high chemical bond strength due to the hydrophilic nature of PVA fibre and high
alkali resistance characteristic. The physical tensile strength and elastic modulus of PVA fibre
are 1600 MPa and 40 GPa, respectively. The failure process in a composite material when a
crack propagates is complex and involves matrix cracking. The bonding strength between fibre
and matrix is to be considered as a source of energy dissipation. Thus, a single fibre pull-out
test was conducted with OPC paste matrix (w/c = 0.3) and AAFA paste matrix (l/s = 0.6)
using PVA fibre, as illustrated in Figure 6.12. Due to the limitation of the mechanical testing
machine, only failure force was captured. The embedded length Ld of fibre is around 4mm,
which is half of the total length, and diameter of fibre df was 0.038mm. Assuming uniform
bonding, the maximum interfacial bonding strength τmax can be expressed as:
τmax =PmaxπdfLd
(6.3.5)
where Pmax is a maximum pull-out force. With equation (6.3.5), the maximum interfacial
bonding strength τmax was determined as shown in Table 6.5. The average elastic modulus of
OPC and AAFA were obtained using indentation modulus results as presented in Chapters 4
and 5. Table 6.6 lists the parameters for the FEM pull-out of OPC and AAFA. Table 6.7 shows
a comparison of the maximum pull-out force between analytical and experimental results which
are good agreement.
Figure 6.12: Schematic of single fibre pull-out test
135
Table 6.5: Maximum interfacial bonding strength
TestOPC AAFA
Pmax (N) τmax (MPa) Pmax (N) τmax (MPa)
1 0.48 1.01 0.53 1.112 0.59 1.24 0.38 0.83 0.44 0.92 0.55 1.154 0.48 1.01 0.61 1.285 0.47 0.98 0.58 1.216 0.48 1.01 0.54 1.137 0.55 1.15 0.53 1.118 0.43 0.9 0.57 1.199 0.44 0.92 0.58 1.2110 0.47 0.98 0.41 0.86
Mean 0.48 1.01 0.53 1.11Std 0.05 0.11 0.07 0.16
Table 6.6: Parameters for analytical modelling of fibre pull-out
OPC matrix
Elastic modulus (MPa) 21218Poisson’s ratio 0.24
AAFA matrix
Elastic modulus (MPa) 28011Poisson’s ratio 0.24
PVA fibre
Elastic modulus (MPa) 40000Tensile strength (MPa) 1600
Poisson’s ratio 0.3Diameter (mm) 0.038
Interface
Embedded length (mm) 4mmMatrix to fibre size ratio 263.15
τmax (MPa) for OPC matrix 1.01τmax (MPa) ofr AAFA matrix 1.11
Asymptotic frictional stress -Reference slip (mm) (Zhan and Meschke, 2014) 0.25
Table 6.7: Maximum pull-out force between analytical and experimental results
FEM Experimental Ratio
Maximum Pull-outForce(N)
OPC 0.482 0.480 1.001
AAFA 0.530 0.530 1.00
136
6.4 High-Performance Fibre Reinforced Cementitious Com-
posite
An experimental study was conducted to understand HPFRCC in AAFA and OPC matrices
which led to the development of HPFRCC. In this research, Class F (low calcium) fly ash from
Australia and Local general purpose (GP) Portland cement according to Australia Standard
(AS 3972) (Australia Standard, 2010) were used for the cementitious matrix. The summary of
the chemical compositions of OPC and fly ash is presented in Table 6.8, XRD pattern of fly
ash is shown in Figure 6.13. The specimens were cast in 25mm cubic mould for compressive
strength test, the three prismatic plate specimens of 160 × 40 × 40 mm in dimensions for flex-
ural performance. AAFA specimens were cured 24 hours at 60◦C, which is a common curing
temperature for AAC (Joseph, 2011; Hardijito and Rangan, 2005; Rangan, 2008). After that,
the specimens were placed in a curing room at 23◦C ± 3 until testing. Compressive strength
test was conducted on 7, 14 and 28 curing days using 200 N/s of loading rate. The flexural per-
formance was conducted according to ASTM ASTM Standard C1609 (2012) with 0.05mm/min
of loading rate on each specimen. A schematic installation for the flexural performance is shown
in Figure 6.14.
The selected mixing proportion is the process of choosing suitable fibre volume fraction of AAFA
and OPC mixture as shown in Table 6.9. Series A is composite AAFA mixture, in groups A
and B notate that mixtures were without silica fume and with silica fume, respectively. Series
P is composite OPC mixtures, in group A and B design that mixtures were without fly ash and
with fly ash, respectively.
Table 6.8: Chemical composition of OPC (type I) and low calcium fly ash (wt. %)
SiO2 Al2O3 CaO Fe2O3 K2O MgO SO3
OPC 21.1 4.7 63.6 2.7 - 2.6 2.5Fly ash 65.9 24.0 1.59 2.87 1.44 - -
Figure 6.13: XRD pattern of fly ash
137
Figure 6.14: A schematic of flexural performance
Table 6.9: Composition of mix proportions
(a) AAFA composites
Series Group Index FA SF l/s SP Vf
A
A
AA1 1 - 0.5 0.02 -AA2 1 - 0.5 0.02 0.5%AA3 1 - 0.5 0.02 1.0%AA4 1 - 0.5 0.02 2.0%
B
AB1 1 0.2 0.5 0.02 0.5%AB2 1 0.2 0.5 0.02 1.0%AB3 1 0.2 0.5 0.02 2.0%AB4 1 0.2 0.5 0.02 -
(b) OPC composites
Series Group Index OPC FA w/c Vf
AA
PA1 1 - 0.4 1.0%PA2 1 - 0.4 2.0%
BPB1 1 0.2 0.4 1.0%PB2 1 0.2 0.4 2.0%
138
6.4.1 Compressive Strength
Compressive Strength Development
Figure 6.15 and Figure 6.16 show the average compressive strength development between 7 to 28
days of curing ages in the different composites. The average compressive strength development
results were obtained from the average of six specimens per mixture. It can be seen that
compressive strength of AAFA and OPC composites generally decreased by fibre volume ratio.
Also, it was observed that compressive strength development is not significantly increased by
fibre volume ratio except A2, which exhibited a high rate of compressive strength development
between 7 to 14 days. The test results indicate that compressive strength development is not
significantly affected by the fibre volume ratio in both AAFA and OPC composites. The test
results are presented in Table 6.10.
Figure 6.15: Compressive strength development of AAFA composites,Series A
139
Figure 6.16: Compressive strength development of OPC composites, Series P
Table 6.10: Compressive strength development (MPa)
Series Index7 days 14 days 28 days
fck,cubic,7 StD fck,cubic,14 StD fck,cubic StD
A
AA1 72.53 9.55 75.45 4.61 75.62 9.91AA2 54.09 10.14 73.99 3.35 71.45 5.47AA3 58.41 5.69 66.98 3.68 70.62 3.37AA4 56.34 3.73 57.88 3.16 61.01 1.69AB1 57.5 6.73 57.39 6.63 64.16 4.76AB2 54.05 4.85 58.07 3.44 61.7 3.33AB3 54.06 2.54 55.7 1.02 57.57 4.26AB4 49.46 2.45 54.5 2.21 56.42 4.10
P
PA1 41.64 6.93 43.86 5.3 44.84 1.79PA2 34.54 2.99 37.41 3.02 37.98 1.92PB1 28.77 2.09 30.37 2.81 34.28 1.41PB2 30.87 2.26 31.82 2.05 33.04 0.79
140
6.4.1.1 Compressive Failure Mode
The behaviour and ultimate compressive failure mode of composites are shown in Figure 6.17
and Figure 6.18. It is known that PVA fibre matrix can exhibit ductile behaviour after reaching
its ultimate compressive strength because of the transverse confinement effect of the PVA fibre
while normal AAFA mixtures without PVA fibres (AA1, AB1) present a significant decrease
in stress after reaching the ultimate compressive strength. However, OPC composites have
more ductile behaviour after reaching its ultimate compressive strength compared to AAFA
composites. The effect of the fibre volumetric ratio on compressive strength and strain capacity
of AAFA composites can be seen in Figure 6.19. Figure 6.19(a) shows the effect of the fibre
volume ratio on the strain at the ultimate compressive strength. It can be seen that the
compressive strain is not significantly affected by the fibre volume ratio. Also, the compressive
strain corresponding to the compressive strength is not meaningfully affected. As shown in
Figure 6.19(b), compressive strength generally decreases with increasing fibre volume ratio in
Figure 6.17: Typical stress-strain curves of AAFA composites, Series A
141
Figure 6.18: Typical stress-strain curve of OPC composites, Series P
the AAFA composites. Silica fume was added to achieve lower compressive strength in AAFA
composites. As discussed in Chapter 5 silica fume in AAFA matrix contributes to a significant
decrease in the compressive strength due to a decrease in the cohesion of the reaction products.
The effect of the fibre volume ratio on compressive strain capacity of OPC composites was
also observed, as summarised in Table 6.11. The results show that fly ash contents of OPC
composites tend to decrease the compressive strain capacity.
Table 6.11: Compressive strain and strength test results
Series IndexStrain at Ultimate Compressive
Compressive strength strength
Strain (%) StD fck,cubic (MPa) StD
A
AA1 2.189 0.174 75.62 9.91AA2 2.226 0.102 71.45 5.47AA3 2.011 0.068 70.62 3.37AA4 1.866 0.125 61.01 1.69AB1 1.896 0.108 64.16 4.76AB2 2.000 0.159 61.70 3.33AB3 1.883 0.173 57.57 4.26AB4 2.020 0.391 56.42 4.10
B
PA1 2.401 0.276 44.48 1.79PA2 2.268 0.371 37.98 1.92PB1 1.578 0.177 34.28 1.41PB2 2.165 0.161 33.04 0.79
142
Figure 6.19: The effect of fibre volume ratio on strain and compressive strength, Series A
143
Modulus of Elasticity
According to Euro Standard 1992-1 (de Normalisation, 2015), the static modulus of elasticity
can be determined as:
Ecm = 22
(fcm10
)0.3
(6.4.1)
where fcm is the mean value of cylinder compressive strength at 28 days in MPa. The relation-
ship between fcm and cubic compressive strength (fck,cubic) can be approximated by:
fcm = 0.4381f1.135ck,cubic + 11.123 (6.4.2)
Similarly, the definition of elastic modulus from the compressive strength is (de Normalisation,
2015):
Eexp,cm =0.4fcmε0.4c
(6.4.3)
where ε0.4c is the compressive strain corresponding to the 40% of fcm. Using, Equations (6.4.1)
and (6.4.3), the average modulus of elasticity of each mix was determined, as shown in Table
6.12. The results of the modulus of elasticity from the experimental and Euro standard 1992
are different ratio in range of 9.5 to 10 for Series A and 11 to 14 for Series B. The reason of
these different could be due to the size effect of the specimens (Kim and Yi, 2002).
Table 6.12: Modulus of Elasticity
Series IndexEN 1992 Experimental Ratio
Ecm (GPa) Eexp,cm (GPa) Ecm/Eexp,cm
A
AA1 39.21 3.97 9.880AA2 38.71 3.89 9.950AA3 38.62 4.00 9.660AA4 37.42 3.98 9.400AB1 37.84 3.92 9.650AB2 37.51 3.79 9.900AB3 36.80 3.64 10.110AB4 36.57 3.51 10.420
B
PA1 33.99 2.74 12.410PA2 32.92 2.35 14.010PA3 32.27 2.90 11.130PA4 32.04 2.46 13.020
144
6.5 Flexural Performance
The flexural behaviour of composites will exhibit deflection-hardening, or softening behaviour
and loading capacity after first cracking. The first cracking point of the composite is defined
as Limit of Proportionality (LOP ), and the maximum equivalent flexural strength point of the
composite is defined as Modulus of Rupture (MOR). The equivalent flexural strength can be
determined by using Equation (6.5.1) with the loading capacity at LOP (PLOP ) and at MOR
(PMOR) provided by ASTM ASTM Standard C1609 (2012).
f =PL
bd2(6.5.1)
where L is the span length, b is the average width, and d is the average depth of specimen. In
addition, flexural toughness index T10 correlates approximately with the total energy absorption
capacity of the beams. T10 can be determined as the total area under the load-deflection curve
up to 10% of the maximum load (Ward and Li, 1991).
Figure 6.20 shows the flexural behaviour of typical AAFA composites in Group of Series A.
The flexural performance of AA1 mix shows typical deflection-softening behaviour, AA2 mix
shows quasi-deflection-softening behaviour, and AA4 mix shows deflection-hardening behaviour.
However for AA3 mix, some of the specimens have complex behaviour, which are deflection-
hardening and quasi-deflection-softening behaviour. The maximum loading capacity of AA4
mix is about 74% greater than that of other mixes and the deflection capacity of AA4 mix is
also greater than that of AA1, AA2 and AA3 mixes. Similarly, Figure 6.21 shows the flexural
behaviour of typical AAFA composites in Group B of Series A. The flexural performance of
AB1, AB2 and AB3 mixes shows typical deflection-softening behaviour, and AB4 mix presents
deflection-hardening behaviour. Maximum loading and deflection capacity of mix AB4 is around
65% and 85% greater than other mixes, respectively.
Similarly, the flexural behaviour of typical OPC composites is illustrated in Figure 6.22. The
flexural results of all mixtures show deflection-hardening behaviour. Maximum loading capacity
of PA2 has around 60% higher than other composites. The maximum deflection capacity of
PA2 composite mix is the highest compared to others.
145
Figure 6.20: Flexural behaviour of AAFA composites in Group A of Series A
Figure 6.21: Flexural behaviour of AAFA composites in Group B of Series A
146
Figure 6.22: Flexural behaviour of OPC composite, Series P
As the volume fraction of fibre content of AAFA composites increase from 0% to 2.0%, Figure
6.23 shows the effect of fibre volume fraction on the deflection capacity of different mixtures of
AAFA composites. These results show an increasing trend of deflection capacity at LOP as the
linear relationship, as shown in Figure 6.23(a), and an increasing trend of deflection capacity
at MOR, as the exponential relationship, as shown in Figure 6.23(b). The improvement of
deflection at MOR, in Group A of Series A was observed much higher deflection capacity
compared with Group B of Series A. As shown in Figure 6.24(a), the increase in fibre volume
fraction leads to an increase in flexural strength at LOP as a linear relationship. The increasing
trend of flexural strength at MOR was observed as the exponential relationship as shown in
Figure 6.24(b). The improvement of flexural strength at MOR in Group A of Series A was
obtained much higher flexural strength capacity at MOR while increasing fibre volume fraction.
Toughness (T10) was determined by trapezoidal numeral integration method with MATLAB
(2014). The influence of fibre volume fraction on toughness is presented in Figure 6.25. T10 of
mixtures with different fibre volume fraction ratio shows increasing trend as a power trend in
both Group A and B of Series A. The overall toughness in Group A of Series A is higher than
Group B of Series A was observed.
147
Figure 6.23: Effect of fibre volume fraction on deflection capacity in AAFA composites, Series A
148
Figure 6.24: Effect of fibre volume fraction on flexural strength in AAFA composites, Series A
Figure 6.25: Effect of fibre volume fraction on toughness in AAFA composites, Series A
149
Similarly, Figure 6.26 shows the effect of fibre volume fraction on the deflection capacity of
OPC composites having the volumetric fibre content 1.0 and 2.0%. The results show that
deflection capacity at LOP and MOR is increased by increasing the fibre volume content. It
can be noticed that fly ash content in OPC matrix does not improve the deflection behaviour
of OPC composites. As shown in Figure 6.27, an increase in volumetric fibre fraction indicates
an increase in flexural strength at LOP and MOR. The improvement of the flexural strength
at MOR in Group A of Series A is evident with an increase in fibre volume fraction. In Figure
6.28 shows the results of T10 with different fibre volume fraction ratio it can be seen that fly ash
content in OPC composites leads to an increase in toughness when the fibre volume fraction as
2.0%. The flexural response can be tested AAFA and OPC mixtures are summarised in Table
6.13.
Li et al. (1995) reported that adding fine aggregates in OPC composite can improve the pseudo-
strain hardening behaviours. However in AAFA composite, adding fine aggregates (SF) does
not improve flexural deflection and strength capacity. Adding Fine aggregate (SF) in AAFA
composite could decrease the flexural strength and increase the toughness of matrix containing
2.0% of fibre volume fraction. There is no improvement in the flexural deflection and the
strength capacity of the mixtures containing fly ash.
Figure 6.26: Effect of fibre volume fraction on deflection of OPC composites, Series P
150
Figure 6.27: Effect of fibre volume fraction on flexural strength of OPC composites, Series P
Figure 6.28: Effect of fibre volume fraction on toughness of OPC composites, Series P
151
Table
6.13:
Fle
xura
lb
ehav
iours
Ser
ies
Ind
exL
imit
ofP
rop
orti
onal
ity
(LOP
)M
od
ulu
sof
Ru
ptu
re(M
OR
)T
ough
nes
s
δ LOP
StD
f LOP
StD
δ MOR
StD
f MOR
StD
T10
(kJ)
StD
(mm
)(M
Pa)
(mm
)(M
Pa)
A
AA
10.1
410.
009
1.23
40.
111
0.14
10.
009
1.23
40.
111
--
AA
20.
176
0.0
641.
303
0.25
60.
176
0.06
41.
303
0.25
60.
048
0.01
9A
A3
0.20
30.0
611.
357
0.13
30.
527
0.31
51.
514
0.12
20.
266
0.08
5A
A4
0.24
20.1
031.
397
0.22
22.
647
0.33
45.
214
1.00
73.
438
0.86
5A
B1
0.1
460.
022
1.36
80.
161
0.14
60.
022
1.36
80.
161
--
AB
20.1
630.
028
1.63
10.
190
0.16
30.
028
1.63
10.
190
0.04
40.
013
AB
30.2
100.
076
1.72
50.
184
0.21
00.
076
1.72
50.
184
0.05
20.
013
AB
40.2
270.
111
1.89
00.
266
1.20
00.
380
4.62
50.
839
1.41
20.
401
P
PA
10.
454
0.13
64.
177
0.49
00.
966
0.13
06.
383
0.72
81.
536
0.31
4P
A2
0.59
50.1
195.
315
1.32
21.
860
0.16
911
.154
1.35
25.
073
0.72
5P
B1
0.50
40.
023
4.54
70.
379
0.95
00.
220
6.62
10.
797
1.47
80.
434
PB
20.
546
0.086
6.05
30.
615
1.18
80.
257
9.03
20.
744
2.87
71.
175
152
The tensile and compressive behaviour of a composite material can theoretically present the
flexural performance (Naaman, 1972; Naaman and Reinhardt, 2008; Naaman et al., 1974; So-
ranakom and Mobasher, 2008; Ward and Li, 1991). When the flexural behaviour of the compos-
ite is strongly associated with its tensile behaviour, the strain-hardening behaviour in tension
leads to a deflection-hardening behaviour in flexural (Kim et al., 2011). Therefore, the tensile
strength of the composite is related with the flexural performance of the composite. In this
study, the flexural test was conducted that failure mode of AA4 and AB4 composites has strain-
hardening behaviour and AA3 has a combination of quasi-strain-softening or strain-hardening
behaviour. Also, OPC composites show strain-hardening behaviour.
Based on the theoretical discussion in the previous section, the critical energy release rate (Gc)
and interfacial bond strength (τ) of the composite are important parameters to be considered
in matrix design to achieve the strain-hardening behaviour of the composite. The matrix
properties, such as elastic modulus and fracture toughness, are linked to the composite critical
energy release rate Gc and are affected by several parameters. In this research, the composite
critical energy release rate Gc of AAFA and OPC matrices were found to be 0.010 kJ/m2 and
0.005 kJ/m2, respectively. Base on nanoindentation test as presented in Chapter 3 and 4. In
Figure 6.29 and Figure 6.30, the test results of Series A and P composites are plotted against
the critical fibre volume fraction ratio and the corresponding strain-hardening behaviour of the
composite, i.e., f is snubbing coefficient which is in term of inclining angle between fibre and
matrix. It can be seen that with 2.0% of fibre volume fraction of AAFA composites, AA4 and
AB4 are in the region of strain-hardening, whereas with less than 0.5% of fibre volume fraction
of the AAFA composites, AA2 and AB2 are not in the region of strain-hardening. It can also be
noticed that with 1.0% of fibre volume fraction of AAFA composites, AA3 and AB3, are partly
in the region of strain-hardening and other parts are fall in the region of strain-hardening. This
is consistent with AA3 composites, which shows a combination of strain-hardening and quasi-
strain-hardening behaviour. It can be seen that the fibre volume fraction of OPC composites,
PA and PB, are in the region of strain-hardening. The fact that strain-hardening behaviour
was achieved in AAFA with 2.0% of fibre volume fraction in the matrix, and OPC with 1.0% of
fibre volume fraction in OPC matrix, the experimental results are consistent with the theoretical
estimation.
153
Figure 6.29: Critical volume fraction against interfacial bond strength with AAFA composites,Series A
Figure 6.30: Critical volume fraction against interfacial bond strength with OPC composites, SeriesP
154
6.6 Chapter Summary
The results of the experimental research on the effect of parameters on fibre pull-out test and
strain-hardening behaviour of the AAFA and OPC composite are presented in this Chapter.
Based on the theoretical discussion of HPFRCC and the properties of matrix presented in
Chapter 3 and 4, the following conclusion can be drawn.
- Interfacial bond strength was determined to be in a range of 0.8 to 1.0 MPa in both OPC
and AAFA matrices.
- An increase of fibre diameter and embedded length could increase interfacial bond strength
between the matrix and the fibre.
- Fine material (SF) in AAFA composites is not suitable. It could decrease the flexural
strength and strain capacity of AAFA composites.
- While achieving the strain-hardening behaviour of the composites, the compressive strength
was decreased.
- The composite critical energy release rate Gc in OPC and AAFA matrix was found using
nanoindentation to be approximately 0.01 kJ/m2. Theoretically, it is impossible to reach
strain-hardening behaviour when Gc is more than 0.015 kJ/m2. Thus, it is recommended
Gc should be less than 0.01 kJ/m2. The results of flexural performance show the strain-
hardening behaviour of AAFA and OPC composites with this value of Gc.
155
Chapter 7
Conclusion and Future Research
This research examines properties of cementitious materials such as Ordinary Portland cement
(OPC) and Alkali-activated fly ash cement (AAFA) and high performance fibre reinforced ce-
mentitious composite. With objective of understanding of the basic properties of cementitious
materials, statistical experiment were performed on OPC and AAFA pastes and mortar with
varying parameters. Indentation properties such as modulus, hardness, packing density, vis-
coelastic properties, stress-strain curves and fracture toughness of cementitious materials were
determined using nanoindentation. With these properties, high-performance fibre reinforced
cementitious composite was studied.
7.1 Summary of Main Findings
From statistical analysis such as Taguchi’s design of experiment, Analysis of variance and
regression, the effect of test factors on properties of cementitious materials were determined.
Based on this study, the following conclusion can be drawn:
a) Nanoindentation technology is used to obtain mechanical properties of reaction products
of blended cements and alkali-activated fly ash cement. Statistical analysis tools is success-
fully used for analysis of indentation results. This technology encourages consideration
of small scales examinate to represent the large scale of civil engineering construction
materials. It is associated with scientific benefits such as comparable to conventional
experimental results, assessment of the microporomechanics and viscous properties of re-
action products of blended cement and alkali-activated fly ash cement, and determinants
of properties in nanoscale enable an assessment of the long-term macroscopic behaviour.
b) OPC is the main binder material in civil engineering thus properties of hydration prod-
ucts of OPC paste were determined using nanoindentation. The structure of CSH govern
fundamental properties such as strength, relaxation, creep and fracture behaviour. The
average of modulus (M) and hardness (H) and packing density (η) of low density CSH
are 16.787 GPa, 0.704 GPa and 0.556, respectively. Indentation properties of high density
of CSH are M = 30.481 GPa, H = 1.415 GPa and η = 0.595. Also, stress relaxation
can be negligible in OPC paste and reducing of elastic modulus due to contact relax-
ation modulus of OPC occurs in very short period. The creep behaviour of OPC paste
mainly occurs capillary porosity which tends to increase creep compliance. From indenta-
tion stress-strain curves, strength failure of OPC paste arise in capillary porosity because
small strain capacity was captured compared to other phases and high capillarity porosity
exhibit low compressive strength. Based on statistical analysis on the results of blended
156
cement mixtures, an increase in fly ash content and a decrease of water to cementitious
material ratio lead to a decrease the density of blended cement. The optimal mix design
for density is 20% fly ash, no content of sand to cementitious material ratio, 0.4 of water
to cementitious material ratio, and 0.1% of superplasticiser. For compressive strength,
an increase in fly ash content and sand to cementitious material ratio decrease the com-
pressive strength development. The optimisation of the compressive strength design is
found to be 20% of fly ash, 1.5 of sand to cementitious material ratio, 0.35 of water to
cementitious material ratio, and 0.2% of superplasticiser. For residual strength of blended
cement, increasing in fly ash and superplasticiser improve the overall residual strength.
The optimal mix design for residual strength, 20% of fly ash, 1.5 of cementitious material
ratio, 0.35 of water to cementitious material ratio, and 0.2% of superplasticiser. The
regression model is proposed to estimate the properties of blended cement mixture.
c) From statistical analysis of AAFA mixtures, silica fume is the most adverse impact on the
compressive strength and increase of superplasticiser and liquid to solid ratio contribute
to the decrease in compressive strength. It is found that no significant effect of sand to
cementitious material ratio on compressive strength. The optimum mix of compressive
strength is no silica fume, no sand, and no superplasticiser with liquid to solid ratio
0.6. In term of density, the increase in liquid to solid ratio and superplasticiser dosage
further decreases the density of AAFA. Properties of reaction products are determined
to be understood Druker-Parker strength and Coulomb material models. It is found that
stiffness and cohesion of reaction products of AAFA have an inverse relationship, i.e., an
increase of stiffness leads to a decrease of cohesion. Four main reaction phases viz., N-A-S-
H, Partly-activated slag, Non-activated slag and Non-activated compact glass phases are
identified. It confirms that N-A-S-H phase is the major reaction products of AAFA and
its volume fraction relates to the compressive strength. The volume fraction of N-A-S-H
phase of 0.50 or over and 0.6 of activation degree is recommended to obtain high strength
of AAFA. Also, it is found that stress relaxation and reduction of elastic modulus of AAFA
can be negligible due to short period of reduction of relaxation modulus. Partly-activated
and Non-activated phases are the main phases an increase of creep behaviour of AAFA
due to “block-polymerisation” concept. Liquid to solid ratio has the most effect on creep
behaviour of AAFA and it leads to an increase of creep behaviour. Sand to cementitious
material ratio and superplasticiser have a minor effect on creep behaviour. In term of
fracture toughness, an increase of silica fume leads to an increase of fracture toughness.
The overall fracture toughness of AAFA is 0.474 MPa m1/2.
d) Based on properties of cementitious materials from statistic and indentation test results,
high performance fibre reinforce cementitious composite is studied. In term of fibre pull
out test, interfacial bond strengths is determined in the range of 0.8 to 1.0 MPa for
OPC and AAFA matrix. An increase of fibre diameter and embedded length leads to
an increase of interfacial bond strength between matrix and fibre. Fine aggregate (silica
fume) should be avoided in AAFA matrix due to decrease of flexural strengths and strain
capacity. It confirms that critical energy release rate of OPC and AAFA matrix are
0.010 kJ/m2 and 0.005 kJ/m2. These critical energy release rate can be used for design
high-performance fibre reinforce cementitious composite. To achieve strain-hardening
behaviour, fibre volume fraction is to be more than 1% in OPC, and 2% in AAFA.
157
7.2 Recommendation for Future Research
A better understanding of the properties of cementitious materials such as mechanical and in-
dentation properties is necessary to achieve the required performance. Though this research
provides valuable mechanical and indentation properties using statistical analysis, the outcome
of properties of cementitious material is still limited to the available technique used. Fur-
ther verification using different technique and source materials will complement the findings
presented in this research. Viscoelastic properties are important material characteristics for
civil engineering structure. This results provide a model and indentation properties to predict
viscoelastic properties of OPC and AAFA composite. The same procedure can be used vis-
coelastic properties of other construction materials. Other test methods are recommended for
further verification. High performance reinforced alkali-activated cement composites have been
achieved using PVA fibres. More studies are recommended using different type of fibre and
mixtures.
158
Appendices
Appendix A
This appendix contains the nanoindentation results of Ordinary Portland cement paste.
Table A1: Indentation test of OPC paste
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
33.074 1.731 0.067 174.2 0.972
16.426 0.729 0.018 261.326 0.959
13.99 0.353 0.013 358.844 0.971
17.04 0.239 -0.001 420.592 0.961
14.171 0.397 -0.014 339.852 0.965
13.067 0.467 -0.02 319.797 0.964
38.459 1.365 0.041 187.278 0.971
21.792 0.727 0.02 255.249 0.97
31.104 0.388 0.036 329.782 0.972
17.124 0.368 -0.013 345.814 0.96
6.589 0.172 -0.044 514.782 0.966
21.09 0.438 -0.007 315.91 0.96
18.867 0.443 -0.004 317.475 0.965
42.001 2.451 0.01 148.665 0.972
108.051 8.391 0.015 84.689 0.968
18.21 0.702 -0.001 262.456 0.962
27.672 0.874 0.004 231.626 0.97
11.883 0.45 -0.053 327.443 0.961
12.104 0.334 -0.034 370.976 0.969
11.115 0.195 -0.043 469.898 0.96
12.462 0.184 -0.029 481.824 0.967
12.566 0.425 -0.032 333.587 0.963
12.356 0.263 -0.046 410.048 0.965
16.384 0.573 -0.029 287.41 0.96
22.327 0.588 -0.032 277.083 0.962
16.215 0.478 -0.035 310.622 0.963
33.133 0.852 0.002 229.585 0.962
213.215 17.474 0.033 60.309 0.972
26.558 0.399 -0.068 326.468 0.966
52.961 0.418 -0.003 312.726 0.969
159
8.946 0.238 -0.076 437.513 0.964
26.619 0.36 -0.019 342.003 0.965
11.374 0.206 -0.064 457.253 0.958
12.328 0.355 -0.067 360.378 0.965
16.568 0.415 -0.036 330.647 0.971
12.003 0.27 -0.097 405.571 0.964
53.949 1.865 0.002 159.484 0.968
20.265 0.702 -0.042 259.491 0.96
15.481 0.153 -0.032 521.427 0.967
68.785 9.177 0.003 92.439 0.969
12.349 0.334 -0.064 368.991 0.962
13.99 0.604 -0.039 288.206 0.972
14.1 0.373 -0.015 350.29 0.971
11.921 0.276 -0.033 401.895 0.964
17.62 0.444 -0.021 319.097 0.968
22.269 0.635 -0.04 268.259 0.961
20.067 0.451 -0.042 314.747 0.971
8.644 0.209 -0.094 463.994 0.966
15.158 0.65 -0.062 276.524 0.964
63.959 3.216 0.017 126.631 0.966
143.139 14.904 0.028 68.426 0.969
7.411 0.153 -0.112 536.15 0.964
239.257 17.899 0.027 58.355 0.964
24.151 0.923 -0.038 228.51 0.962
26.952 0.489 -0.021 298.392 0.972
37.651 1.252 0.002 193.92 0.967
26.191 0.926 -0.016 226.361 0.962
19.715 0.714 -0.052 259.879 0.971
29.591 0.511 0.007 290.203 0.966
37.942 1.199 -0.024 197.466 0.969
31.629 0.811 0.001 236.664 0.973
103.127 4.041 0.034 109.931 0.971
14.997 0.259 -0.056 407.513 0.961
7.168 0.181 -0.092 499.726 0.963
13.939 0.579 -0.053 293.013 0.971
25.255 0.673 -0.015 259.375 0.963
13.971 0.309 -0.03 378.105 0.96
10.477 0.233 -0.056 436.979 0.964
64.489 2.698 0.03 135.47 0.971
88.459 2.33 -0.017 139.171 0.966
23.083 0.789 -0.03 244.861 0.964
14.457 0.423 -0.023 330.45 0.966
19.158 0.649 -0.047 269.017 0.959
15.494 0.387 -0.032 341.817 0.967
11.923 0.379 -0.059 351.685 0.965
17.901 0.559 -0.034 288.532 0.963
9.467 0.187 -0.07 482.605 0.958
160
9.538 0.158 -0.091 521.714 0.964
8.971 0.17 -0.075 504.661 0.958
171.068 5.412 0.02 65.214 0.482
14.832 0.438 -0.043 323.831 0.96
19.869 0.356 -0.028 348.71 0.964
14.404 0.289 -0.056 390.262 0.968
41.219 2.162 -0.017 155.553 0.968
320.036 10.325 0.036 67.801 0.965
28.846 1.164 -0.017 205.067 0.966
24.496 0.986 -0.016 222.939 0.967
13.219 0.272 -0.061 403.003 0.969
8.362 0.132 -0.087 571.744 0.968
12.007 0.316 -0.04 380.199 0.97
15.891 0.528 -0.038 298.526 0.963
52.297 2.211 0.024 149.551 0.967
19.428 0.715 -0.006 259.557 0.967
6.238 0.177 -0.118 509.159 0.962
21.962 0.587 -0.047 277.648 0.962
110.011 11.139 0.01 78.249 0.97
36.309 0.68 0.023 253.361 0.971
25.368 0.616 -0.003 270.461 0.971
14.701 0.449 -0.033 322.146 0.966
9.718 0.308 -0.113 390.98 0.97
11.152 0.292 -0.082 392.888 0.958
14.892 0.366 -0.054 350.543 0.965
11.53 0.306 -0.072 384.41 0.959
12.171 0.221 -0.027 442.43 0.96
12.41 0.252 -0.049 417.308 0.964
17.055 0.704 -0.046 263.813 0.96
38.74 0.708 -0.003 247.702 0.971
10.505 0.235 -0.043 436.334 0.969
16.616 0.539 -0.038 295.189 0.966
11.199 0.153 -0.067 524.921 0.96
9.26 0.231 -0.063 441.336 0.962
13.425 0.327 -0.047 370.267 0.962
25.675 0.831 -0.037 237.745 0.968
21.345 0.481 -0.021 303.031 0.96
41.704 1.641 0.007 172.136 0.966
7.386 0.151 -0.087 538.863 0.96
24.179 0.716 -0.013 254.505 0.97
79.862 3.888 0.022 114.452 0.961
21.219 0.303 -0.032 375.409 0.971
18.052 0.589 -0.027 281.758 0.96
22.789 0.71 -0.02 255.274 0.959
5.528 0.07 -0.064 775.834 0.963
10.864 0.26 -0.066 416.634 0.971
15.275 0.365 -0.037 349.343 0.96
161
14.225 0.375 -0.044 349.19 0.971
28.808 0.706 -0.02 251.911 0.965
7.374 0.166 -0.112 518.918 0.967
12.468 0.311 -0.045 379.939 0.96
11.23 0.201 -0.014 463.787 0.962
14.191 0.412 -0.075 335.125 0.969
18.941 0.559 -0.042 288.246 0.971
8.771 0.212 -0.082 461.641 0.967
18.38 0.17 -0.055 491.515 0.961
63.245 3.057 0.013 129.418 0.969
17.667 0.524 -0.059 296.336 0.96
29.298 0.735 -0.026 248.072 0.971
9.711 0.195 -0.085 475.577 0.967
15.325 0.528 -0.037 299.508 0.963
43.962 2.051 -0.001 157.428 0.971
23.143 0.72 -0.022 254.876 0.97
33.095 3.497 -0.007 139.882 0.97
13.509 0.351 -0.026 358.46 0.959
25.147 0.57 -0.024 279.74 0.969
28.73 0.731 -0.012 247.529 0.96
7.756 0.086 -0.227 697.816 0.967
21.549 0.636 -0.039 270.139 0.97
79.983 1.083 0.045 138.056 0.479
3.725 0.062 -0.349 832.771 0.959
37.571 0.193 -0.01 456.747 0.967
10.667 0.195 -0.06 471.423 0.963
4.883 0.124 -0.149 602.86 0.958
31.781 1.098 -0.009 208.404 0.972
11.456 0.391 -0.054 348.607 0.965
9.999 0.263 -0.058 416.348 0.968
11.216 0.26 -0.035 414.2 0.962
21.563 0.602 -0.062 274.754 0.959
48.034 0.959 0.033 213.681 0.969
13.806 0.223 -0.021 439.818 0.97
11.775 0.31 -0.064 384.224 0.97
10.375 0.093 -0.061 666.406 0.967
22.895 0.698 -0.034 257.493 0.963
15.67 0.321 -0.031 369.13 0.96
15.505 0.212 -0.039 447.999 0.971
5.276 0.066 -0.092 794.716 0.959
8.978 0.182 -0.075 492.337 0.969
18.608 1.202 -0.01 215.114 0.964
35.187 3.142 0.002 141.585 0.966
12.514 0.219 -0.112 443.412 0.961
24.955 1.069 -0.019 215.692 0.968
64.564 1.478 0.023 173.184 0.967
17.382 0.577 -0.05 286.407 0.969
162
14.347 0.352 -0.08 356.925 0.962
31.476 1.339 -0.037 191.877 0.962
26.368 0.844 -0.024 234.846 0.961
9.35 0.227 -0.051 445.24 0.964
26.293 0.277 -0.011 386.072 0.96
13.007 0.206 -0.038 454.588 0.959
14.618 0.264 -0.036 405.17 0.965
12.356 0.263 -0.046 410.048 0.965
16.384 0.573 -0.029 287.41 0.96
22.327 0.588 -0.032 277.083 0.962
16.215 0.478 -0.035 310.622 0.963
33.133 0.852 0.002 229.585 0.962
213.215 17.474 0.033 60.309 0.972
26.558 0.399 -0.068 326.468 0.966
52.961 0.418 -0.003 312.726 0.969
8.946 0.238 -0.076 437.513 0.964
26.619 0.36 -0.019 342.003 0.965
11.374 0.206 -0.064 457.253 0.958
12.328 0.355 -0.067 360.378 0.965
16.568 0.415 -0.036 330.647 0.971
12.003 0.27 -0.097 405.571 0.964
53.949 1.865 0.002 159.484 0.968
20.067 0.451 -0.042 314.747 0.971
8.644 0.209 -0.094 463.994 0.966
15.158 0.65 -0.062 276.524 0.964
17.901 0.559 -0.034 288.532 0.963
9.467 0.187 -0.07 482.605 0.958
9.538 0.158 -0.091 521.714 0.964
Appendix B
This appendix contains the nanoindentation results of Alkali-activated fly ash cement.
Table B1: Indentation test of AAFA mix 1
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
5.704 0.091 0.507 688.717 0.971
16.203 0.451 0.505 318.436 0.963
58.588 7.23 0.608 101.607 0.969
73.975 9.24 0.538 90.616 0.974
15.113 0.813 0.456 254.66 0.965
22.489 4.181 0.482 151.609 0.964
35.025 5.37 0.486 125.149 0.962
17.595 0.637 0.416 274.252 0.965
14.962 0.468 0.392 315.691 0.965
163
15.583 0.525 0.373 300.505 0.967
11.498 0.445 0.348 330.411 0.964
14.291 0.616 0.345 284.745 0.967
11.379 0.501 0.298 316.735 0.969
10.983 0.213 0.33 452.874 0.965
32.541 3.128 0.362 144.07 0.963
12.981 1.035 0.312 241.274 0.967
66.244 7.507 0.355 97.567 0.971
10.003 0.374 0.199 360.351 0.969
10.17 0.302 0.248 391.119 0.962
46.893 2.42 0.334 147.097 0.974
37.716 1.859 0.319 166.227 0.967
29.234 1.961 0.298 169.136 0.961
51.216 9.077 0.316 101.91 0.97
51.159 4.88 0.297 115.792 0.972
13.221 0.647 0.224 281.825 0.965
18.985 3.932 0.278 162.671 0.962
32.811 3.961 0.229 135.604 0.963
10.654 0.321 0.184 380.206 0.963
64.88 4.695 0.255 110.996 0.963
10.922 0.478 0.164 324.384 0.97
11.546 0.409 0.161 341.244 0.96
8.549 0.238 0.172 439.631 0.966
28.237 0.86 0.207 232.779 0.971
14.489 2.052 0.219 197.308 0.963
29.53 3.809 0.216 141.03 0.965
52.621 8.321 0.221 102.348 0.972
41.732 3.582 0.215 131.944 0.972
74.724 10.121 0.209 88.551 0.97
13.804 0.675 0.156 275.132 0.96
28.558 2.619 0.18 155.618 0.961
18.465 0.847 0.154 243.607 0.961
67 3.06 0.199 128.458 0.97
49.957 7.726 0.198 105.362 0.971
9.732 0.487 0.123 326.706 0.969
41.785 3.39 0.192 133.389 0.963
23.528 1.011 0.149 222.507 0.973
9.847 0.459 0.073 333.025 0.966
11.896 0.51 0.102 312.051 0.963
11.555 0.325 0.112 376.524 0.968
7.873 0.364 0.082 373.051 0.962
21.323 0.494 0.116 300.352 0.965
11.172 0.438 0.081 332.996 0.96
16.085 1.324 0.143 214.716 0.967
21.405 1.017 0.153 224.444 0.972
14.255 0.567 0.081 293.379 0.963
15.493 0.638 0.101 277.193 0.96
164
13.782 0.579 0.089 292.228 0.964
10.286 0.453 0.045 332.83 0.966
89.355 11.026 0.166 82.612 0.968
33.881 1.848 0.129 169.076 0.967
17.661 0.71 0.074 261.911 0.96
32.409 3.774 0.142 137.743 0.965
38.14 3.295 0.132 136.955 0.961
16.85 0.789 0.09 253.176 0.962
29.689 1.637 0.15 180 0.967
17.75 0.69 0.082 264.976 0.962
37.877 6.391 0.153 118.566 0.962
35.052 1.395 0.12 187.291 0.969
17.913 1.056 0.091 226.739 0.969
18.989 0.878 0.068 240.687 0.97
12.915 0.536 0.065 304.527 0.971
16.508 0.694 0.061 266.16 0.96
13.373 0.734 0.048 268.74 0.965
20.424 2.373 0.092 173.302 0.962
51.285 6.919 0.131 106.563 0.97
25.475 1.71 0.115 182.214 0.972
6.306 0.084 -0.08 706.358 0.956
15.554 1.564 0.079 206.053 0.964
31.877 1.864 0.115 170.332 0.969
86.901 11.511 0.129 82.55 0.968
61.84 7.972 0.121 98.249 0.973
19.8 4.036 0.096 159.463 0.961
19.059 3.404 0.08 165.213 0.96
19.771 0.823 0.054 244.312 0.961
16.288 0.587 0.097 286.698 0.972
17.695 0.824 0.084 247.246 0.96
13.476 0.686 0.02 275.133 0.965
39.549 5.28 0.098 120.946 0.962
14.191 0.584 0.03 290.175 0.963
12.524 0.502 0.024 312.17 0.963
83.372 11.545 0.116 83.512 0.968
22.823 1.598 0.084 189.727 0.97
11.876 0.232 0.007 432.94 0.957
20.577 2.963 0.105 164.918 0.962
57.653 6.994 0.097 102.995 0.973
41.455 7.281 0.111 112.775 0.963
20.256 0.665 0.077 266.232 0.967
32.879 5.314 0.1 128.336 0.967
62.502 8.915 0.077 95.654 0.969
40.838 2.159 0.076 155.77 0.968
28.312 0.678 -0.081 256.181 0.961
8.371 0.128 -0.15 577.917 0.961
9.544 0.321 -0.15 384.198 0.964
165
19.071 0.559 -0.133 286.315 0.959
9.187 0.407 -0.162 351.491 0.967
18.287 0.82 -0.127 246.887 0.96
10.065 0.257 -0.179 420.25 0.967
54.82 6.093 -0.107 107.613 0.971
19.424 0.794 -0.096 248.523 0.963
10.585 0.727 -0.151 280.673 0.97
46.839 5.954 -0.091 112.846 0.971
8.945 0.396 -0.185 355.823 0.963
5.665 0.263 -0.254 439.899 0.964
6.699 0.414 -0.208 364.545 0.963
8.966 0.464 -0.203 335.776 0.967
14.219 0.321 -0.179 371.863 0.963
7.147 0.371 -0.212 374.483 0.96
19.353 0.79 -0.132 248.851 0.962
39.918 5.642 -0.115 119.537 0.97
11.511 0.466 -0.189 323.854 0.96
20.983 2.828 -0.141 165.662 0.964
7.168 0.275 -0.253 420.009 0.963
14.794 0.951 -0.149 241.182 0.96
10.627 0.636 -0.202 293.192 0.969
13.454 0.321 -0.188 373.005 0.963
10.161 0.538 -0.211 313.278 0.97
7.181 0.328 -0.232 393.617 0.968
3.45 0.081 -0.415 742.565 0.955
8.442 0.429 -0.228 347.608 0.962
56.197 8.192 -0.142 100.215 0.966
12.991 0.557 -0.216 298.56 0.963
58.295 6.962 -0.147 102.436 0.966
17.275 0.799 -0.198 250.836 0.959
13.72 0.437 -0.217 327.724 0.966
10.397 0.342 -0.238 370.367 0.96
35.333 2.637 -0.147 148.673 0.961
15.198 0.789 -0.19 257.65 0.969
54.871 7.037 -0.151 104.219 0.971
9.014 0.521 -0.233 320.871 0.96
7.713 0.236 -0.292 444.888 0.968
24.498 2.49 -0.149 163.907 0.967
10.037 0.499 -0.235 320.699 0.959
52.274 5.93 -0.147 109.642 0.971
91.062 9.486 -0.14 84.932 0.965
71.518 2.725 -0.171 133.454 0.972
47.205 6.056 -0.172 112.209 0.971
32.056 3.787 -0.154 138.288 0.969
5.768 0.17 -0.372 522.139 0.964
26.254 2.316 -0.163 164.012 0.96
63.736 7.203 -0.187 99.415 0.969
166
9.866 0.406 -0.288 348.612 0.964
33.525 3.882 -0.209 135.789 0.967
16.296 0.669 -0.224 270.313 0.959
12.468 0.832 -0.179 260.613 0.967
31.683 4.418 -0.18 134.322 0.969
12.452 0.433 -0.246 331.718 0.965
52.881 7.163 -0.18 104.959 0.971
25.954 3.577 -0.176 148.2 0.963
8.208 0.378 -0.293 366.883 0.968
23.716 1.425 -0.217 195.079 0.963
139.765 7.106 -0.156 85.716 0.968
34.183 2.403 -0.204 154.217 0.964
12.224 0.451 -0.261 326.649 0.963
50.104 2.689 -0.199 139.765 0.966
23.08 0.589 -0.239 275.883 0.959
13.718 0.607 -0.256 287.407 0.965
16.724 0.643 -0.238 275.057 0.967
7.105 0.377 -0.375 373.003 0.961
25.728 3.344 -0.235 150.421 0.96
9.484 0.436 -0.288 340.847 0.965
14.675 0.553 -0.267 295.123 0.962
13.668 0.813 -0.365 257.549 0.958
10.101 0.415 -0.291 344.345 0.962
8.099 0.472 -0.315 337.502 0.959
57.207 3.992 -0.215 119.872 0.969
31.063 1.624 -0.181 179.69 0.971
13.384 0.389 -0.306 344.001 0.964
46.9 4.241 -0.214 122.624 0.971
79.523 10.062 -0.19 86.726 0.962
73.954 7.35 -0.204 95.372 0.969
10.848 0.455 -0.31 328.698 0.957
21.023 0.427 -0.252 320.873 0.968
21.496 1.467 -0.302 196.502 0.964
36.271 4.368 -0.186 129.531 0.969
8.674 0.345 -0.336 377.207 0.966
10.147 0.504 -0.293 320.179 0.965
29.275 1.335 -0.224 194.062 0.965
14.61 0.625 -0.28 282.83 0.971
51.738 6.186 -0.218 108.891 0.971
14.636 0.492 -0.312 309.036 0.958
27.925 3.254 -0.271 147.927 0.96
10 0.439 -0.319 336.83 0.96
9.541 0.438 -0.337 339.262 0.961
23.334 1.586 -0.237 189.521 0.971
20.272 1.898 -0.229 183.428 0.959
18.315 1.057 -0.313 226.001 0.97
32.983 2.586 -0.222 151.853 0.965
167
2.135 0.011 -0.52 1937.884 0.946
5.809 0.048 -0.626 926.94 0.957
Table B2: Indentation test of AAFA mix 2
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
70.057 7.909 0.013 95.056 0.972
11.786 0.297 -0.06 389.661 0.962
5.084 0.106 -0.267 645.66 0.962
16.043 0.98 -0.079 235.638 0.96
18.313 0.996 -0.041 229.692 0.96
15.765 0.609 -0.085 282.547 0.965
73.772 7.817 0.004 94.095 0.972
19.343 1.082 -0.049 221.121 0.96
15.342 0.816 -0.066 253.515 0.964
38.128 4.976 0.075 123.596 0.96
12.606 0.7 -0.074 276.376 0.97
65.71 5.683 -0.014 104.961 0.97
14.653 0.758 -0.054 263.008 0.97
13.595 0.704 -0.062 271.729 0.962
27.637 1.162 -0.02 205.749 0.962
37.386 5.28 0.022 123.372 0.968
14.245 0.631 -0.032 282.134 0.967
13.401 0.597 -0.082 289.849 0.965
18.301 0.538 -0.055 293.602 0.97
13.972 0.728 -0.047 268.248 0.968
72.366 7.414 0.065 95.264 0.962
11.602 0.5 -0.069 315.343 0.963
10.292 0.427 -0.086 340.793 0.969
74.14 7.548 0.038 94.728 0.97
6.819 0.143 -0.202 556.557 0.965
6.214 0.064 0.062 811.325 0.972
13.131 0.662 -0.013 279.734 0.965
50.576 5.737 0.067 111.364 0.969
14.486 2.882 0.037 187.816 0.97
87.062 8.093 0.057 89.296 0.963
83.369 7.523 0.07 92.177 0.968
3.114 0.024 -0.101 1313.436 0.968
66.462 8.057 0.076 95.942 0.97
12.935 0.576 -0.017 294.366 0.959
79.539 7.56 0.059 92.646 0.96
19.814 1.676 0.034 192.194 0.971
11.496 0.484 -0.011 319.112 0.96
38.655 1.61 0.062 174.843 0.966
168
22.034 0.555 0.021 285.379 0.968
17.307 0.812 0.037 249.877 0.964
18.684 0.764 0.001 253.265 0.962
52.019 6.913 0.106 106.28 0.972
61.23 7.569 0.053 98.983 0.961
17.967 2.274 0.08 182.136 0.972
17.996 1.246 0.011 214.517 0.971
15.161 0.799 0.032 255.883 0.964
102.67 14.716 0.113 75.09 0.968
13.692 0.791 0.028 261.613 0.971
78.788 7.387 0.105 93.931 0.971
14.639 0.797 0.034 257.125 0.962
60.201 10.004 0.102 94.982 0.969
8.154 0.127 -0.058 581.488 0.967
9.924 0.28 0.061 406.578 0.971
12.608 0.558 0.033 300.662 0.97
24.817 1.602 0.112 186.667 0.969
92.827 9.685 0.122 84.396 0.971
13.561 0.676 0.017 277.269 0.971
16.243 0.407 0.029 333.728 0.97
34.818 3.647 0.09 136.797 0.971
11.158 0.396 0.038 348.436 0.971
8.581 0.519 -0.031 324.448 0.963
21.326 1.859 0.032 182.583 0.96
12.155 0.228 0.085 437.201 0.964
10.103 0.351 0.029 368.632 0.965
113.945 7.747 0.08 85.753 0.963
15.313 0.601 0.019 284.123 0.96
76.169 6.183 0.065 99.4 0.97
12.301 0.277 0.038 399.364 0.959
44.648 5.007 0.071 118.9 0.971
83.778 9.604 0.087 86.83 0.972
12.722 0.762 0.006 266.594 0.96
7.746 0.266 -0.041 422.987 0.964
10.879 0.676 0.024 284.909 0.961
12.767 0.621 0.033 288.16 0.97
21.25 1.226 0.065 209.498 0.967
14.928 0.738 0.056 263.803 0.963
14.039 0.722 0.044 268.102 0.963
11.175 0.379 -0.031 352.886 0.961
44.268 4.016 0.036 126.059 0.971
9.176 0.42 0.002 346.608 0.962
33.524 2.146 0.114 161.064 0.969
13.082 0.718 0.02 272.466 0.97
12.078 0.527 0.011 307.195 0.961
18.791 0.459 0.064 312.219 0.963
14.994 0.877 0.082 247.648 0.961
169
12.204 0.524 -0.013 307.986 0.963
13.611 0.655 0.015 280.073 0.97
18.095 1.012 0.023 229.115 0.963
8.915 1.037 0.051 262.787 0.964
37.562 2.701 0.009 145.786 0.96
12.76 0.52 0.008 308.253 0.97
13.74 0.7 0.035 273.064 0.97
12.793 0.663 0.014 279.735 0.959
12.452 0.641 0.014 285.003 0.965
9.363 0.272 -0.036 411.844 0.963
7.573 0.279 -0.019 415.918 0.965
62.457 2.179 0.138 148.011 0.973
14.331 0.772 0.023 262.171 0.971
12.457 0.543 0.041 302.423 0.96
36.815 1.921 0.059 165.097 0.97
15.075 0.951 -0.01 241.272 0.966
13.799 1.423 0.037 217.563 0.966
10.562 0.6 -0.022 297.963 0.959
12.604 0.589 -0.026 293.432 0.964
70.298 5.192 0.058 106.026 0.964
35.463 3.198 0.043 140.297 0.961
21.325 1.199 0.013 211.607 0.972
56.787 8.975 0.053 98.482 0.969
48.059 7.123 0.054 108.176 0.971
13.015 1.043 0.012 241.074 0.97
43.904 5.984 0.045 114.779 0.969
15.888 0.912 -0.012 242.081 0.961
30.716 1.72 0.039 176.371 0.971
9.925 0.553 -0.052 309.589 0.959
18.173 0.649 -0.035 270.707 0.96
7.33 0.345 -0.082 385.776 0.97
56.386 6.259 0.05 106.242 0.972
15.525 0.638 0.005 277.02 0.96
8.477 0.447 -0.032 342.921 0.965
35.99 2.77 0.033 145.838 0.96
3.242 0.105 -0.216 671.705 0.968
68.195 7.354 0.031 97.252 0.969
66.481 3.499 0.039 122.626 0.973
11.849 0.638 -0.031 288.19 0.969
11.8 0.899 -0.002 256.379 0.964
15.866 1.058 -0.012 231.249 0.97
9.645 0.562 -0.046 310.799 0.969
39.951 1.906 0.041 163.902 0.972
13.815 0.845 -0.012 254.428 0.964
22.247 1.305 0.013 204.068 0.972
33.433 4.371 0.035 132.611 0.971
10.935 1.535 0.01 227.766 0.965
170
12.449 0.619 -0.036 288.174 0.961
14.473 0.429 0 328.73 0.966
10.165 0.536 -0.053 311.944 0.959
37.529 3.693 0.041 134.033 0.973
14.409 0.71 -0.018 269.988 0.971
15.304 0.654 -0.019 276.559 0.972
19.946 1.008 -0.001 225.89 0.96
18.022 1.027 0.017 228.751 0.969
9.38 0.548 -0.049 313.954 0.964
10.766 0.581 -0.044 301.227 0.963
8.869 0.414 -0.044 350.526 0.965
10.478 0.56 -0.023 307.619 0.971
21.439 0.979 0.014 226.357 0.961
8.161 0.37 -0.053 368.821 0.961
11.432 0.57 -0.017 300.349 0.961
9.93 0.544 -0.021 311.623 0.961
62.457 7.207 0.043 100.009 0.97
36.013 5.051 0.023 125.875 0.969
23.295 0.744 0.038 251.052 0.968
8.844 0.445 -0.046 341.568 0.965
12.888 0.567 -0.03 296.917 0.964
15.29 0.988 -0.011 237.898 0.968
10.242 0.482 -0.03 323.994 0.959
64.098 7.306 0.057 99.028 0.971
46.246 6.329 0.035 111.962 0.972
20.875 1.812 -0.006 185.185 0.964
8.892 0.488 -0.068 329.854 0.965
33.728 5.077 0.033 127.886 0.962
10.353 0.517 -0.054 316.097 0.963
11.706 0.487 -0.027 317.618 0.959
19 0.547 -0.027 289.863 0.964
8.97 0.267 -0.081 417.503 0.968
13.32 0.682 -0.01 276.847 0.97
9.871 0.472 -0.032 328.21 0.959
14.96 0.868 -0.009 248.401 0.959
30.652 0.912 0.047 225.653 0.971
16.737 1.654 0.033 200.397 0.972
33.771 5.69 0.026 125.477 0.962
28.379 2.773 0.03 154.263 0.971
63.072 5.8 0.035 105.465 0.973
27.87 0.665 0.017 259.334 0.966
49.755 5.728 0.035 112.048 0.972
8.307 0.371 -0.059 367.687 0.961
9.602 0.381 -0.07 357.797 0.962
18.013 1.053 -0.005 226.815 0.969
9.67 0.479 -0.055 327.703 0.961
26.446 2.693 0.009 157.415 0.963
171
22.101 1.709 0.02 185.869 0.96
15.089 0.663 -0.022 273.732 0.959
12.021 0.524 -0.025 308.411 0.963
49.758 6.551 0.043 108.531 0.967
12.219 0.546 -0.032 304.246 0.971
14.34 0.658 -0.018 277.363 0.966
17.747 1.031 -0.008 229.143 0.97
96.322 8.689 0.064 85.572 0.961
17.267 1.07 -0.009 227.159 0.97
14.565 0.81 0.007 256.009 0.963
9.536 0.527 -0.045 317.524 0.963
20.322 1.507 0.004 197.504 0.971
17.719 1.063 -0.003 226.793 0.97
24.99 1.147 0.014 210.043 0.969
13.878 0.694 -0.017 272.373 0.962
13.314 0.66 -0.028 279.284 0.964
16.996 0.931 -0.007 237.782 0.96
11.956 0.583 -0.037 297.632 0.97
40.326 5.173 0.066 121.267 0.97
13.75 0.728 -0.025 268.603 0.966
Table B3: Indentation test of AAFA mix 3
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
35.431 7.804 0.302 118.979 0.967
39.971 1.613 0.367 174.523 0.97
33.502 2.328 0.342 156.291 0.963
24.274 1.229 0.408 205.169 0.966
12.984 1.318 0.336 225.857 0.971
13.808 0.805 0.262 258.374 0.961
5.534 0.116 0.087 619.297 0.969
30.513 2.214 0.324 162.24 0.971
11.591 0.585 0.28 297.969 0.967
12.91 1.028 0.326 242.787 0.972
49.166 3.989 0.356 122.987 0.962
63.594 6.82 0.339 100.991 0.973
38.36 3.918 0.32 131.157 0.971
12.799 0.39 0.243 346.331 0.971
19.485 1.146 0.302 217.546 0.969
74.815 10.449 0.359 87.764 0.965
82.589 8.796 0.321 88.777 0.969
35.887 1.905 0.402 166.343 0.972
74.261 6.877 0.302 96.897 0.968
17.718 2.383 0.27 180.674 0.968
172
30.651 3.949 0.279 138.299 0.963
96.834 11.528 0.308 79.809 0.962
31.078 1.989 0.299 166.738 0.963
85.463 10.477 0.267 84.36 0.964
79.529 8.029 0.284 91.507 0.966
18.627 2.695 0.254 173.846 0.969
12.072 0.78 0.171 266.916 0.961
55.114 10.135 0.247 97.651 0.967
22.937 0.952 0.231 227.97 0.969
32.48 3.038 0.214 145.885 0.971
10.262 0.449 0.145 333.112 0.96
41.233 4.448 0.236 124.863 0.972
5.862 0.062 -0.022 823.32 0.967
5.606 0.137 0.003 573.802 0.966
33.601 5.376 0.218 127.267 0.969
47.307 5.166 0.213 116.158 0.969
51.871 4.088 0.17 121.394 0.972
10.915 0.412 0.084 342.69 0.963
49.765 6.563 0.177 108.662 0.97
23.014 1.214 0.179 208.207 0.971
11.515 0.467 0.038 323.786 0.961
21.881 1.118 0.089 215.945 0.97
9.301 0.365 0.04 365.452 0.964
39.308 4.641 0.1 124.997 0.969
44.571 3.292 0.122 133.48 0.971
13.265 0.578 0.06 293.138 0.959
69.814 6.604 0.071 99.249 0.967
22.473 1.474 0.056 195.443 0.972
59.729 7.758 0.126 99.592 0.969
29.13 2.487 0.158 157.258 0.962
8.983 0.286 0.018 404.974 0.963
8.501 0.338 0.053 380.163 0.961
19.488 1.949 0.158 184.873 0.97
109.337 19.933 0.182 70.355 0.968
77.668 10.338 0.156 86.968 0.965
34.871 4.911 0.194 127.989 0.971
19.498 1.803 0.163 188.625 0.968
43.405 3.86 0.188 128.13 0.973
10.755 0.47 0.106 326.639 0.967
83.253 9.702 0.168 86.442 0.965
25.52 1.401 0.124 193.889 0.961
54.337 7.026 0.15 104.332 0.968
43.872 6.066 0.218 114.682 0.972
70.269 5.411 0.176 105.211 0.973
17.357 1.717 0.113 196.403 0.969
11.458 0.44 0.064 331.537 0.961
6.904 0.251 0.037 438.46 0.967
173
11.034 0.532 0.086 310.095 0.963
39.541 3.944 0.132 129.403 0.962
18.189 0.904 0.117 237.979 0.959
31.138 4.36 0.101 135.329 0.968
47.004 6.53 0.141 110.757 0.972
15.518 0.313 0.06 374.158 0.964
12.541 1.062 -0.023 240.709 0.963
20.083 1.756 0.126 188.435 0.964
11.323 0.876 0.166 261.551 0.972
30.641 3.568 0.073 141.644 0.965
19.92 2.435 0.131 173.246 0.96
82.665 6.213 0.191 97.314 0.963
5.665 0.136 0.002 574.172 0.96
30.332 2.738 0.069 152.535 0.972
51.755 1.602 0.131 170.805 0.973
84.298 9.782 0.185 85.942 0.964
20.334 0.403 0.076 329.461 0.965
20.219 1.673 0.039 191.082 0.967
9.261 0.196 0.021 473.853 0.962
3.581 0.032 -0.284 1130.572 0.959
3.355 0.029 -0.389 1189.193 0.951
95.944 11.162 0.127 81.008 0.972
15.021 1.293 0.04 219.45 0.969
5.571 0.114 -0.129 620.271 0.962
11.898 0.281 0.01 399.007 0.963
2.309 0.024 -0.576 1308.335 0.952
3.608 0.033 -0.467 1122.277 0.958
11.373 0.431 0.038 336.426 0.971
74.32 8.721 0.078 91.587 0.973
3.83 0.095 -0.291 686.706 0.956
59.449 10.962 0.144 94.104 0.967
9.452 0.338 0 375.827 0.96
14.03 1.499 0 213.309 0.961
20.597 0.71 0.081 258.267 0.964
53.261 6.77 0.072 105.739 0.968
22.249 2.467 0.11 167.698 0.96
14.551 0.866 -0.031 250.53 0.966
72.949 6.56 0.026 98.454 0.967
48.8 5.923 0.05 111.783 0.972
16.945 0.683 0.095 267.903 0.966
6.808 0.362 -0.078 380.869 0.961
13.565 0.733 0.059 269.334 0.972
9.743 0.348 -0.042 371.305 0.966
24.669 1.312 -0.004 200.598 0.972
117.801 16.296 0.063 70.591 0.964
4.76 0.164 -0.113 538.18 0.962
5.52 0.18 -0.082 510.863 0.96
174
5.703 0.08 -0.187 726.287 0.962
74.165 7.761 0.048 93.582 0.961
83.134 7.332 0.051 92.835 0.967
59.587 8.414 0.029 98.223 0.972
24.561 2.941 -0.002 157.363 0.968
14.465 1.422 -0.016 215.809 0.971
161.419 39.931 0.083 57.559 0.968
44.154 5.473 0.043 116.731 0.969
4.091 0.033 -0.291 1104.595 0.957
45.568 5.322 0.042 116.46 0.969
51.511 4.174 0.032 120.217 0.962
16.823 0.341 -0.045 359.524 0.97
12.494 1.036 0.003 242.32 0.96
21.802 0.811 0.014 243.136 0.961
42.035 8.169 0.032 111.024 0.971
10.909 0.345 -0.044 367.63 0.96
13.635 0.849 -0.107 253.736 0.958
11.801 0.781 -0.095 267.588 0.959
5.437 0.118 -0.137 612.275 0.958
92.704 9.763 0.046 83.982 0.965
23.592 1.638 -0.008 187.24 0.972
43.787 2.75 0.05 141.277 0.962
17.292 0.899 -0.034 240.695 0.964
9.569 0.295 -0.051 396.732 0.959
8.264 0.217 -0.121 457.229 0.961
18.399 0.956 0.035 233.022 0.961
15.829 0.589 0.007 285.238 0.96
15.454 0.614 -0.023 281.49 0.961
28.77 1.518 0.022 186.116 0.971
12.21 0.225 -0.026 440.185 0.968
35.495 3.006 0.048 143.55 0.971
6.556 0.233 -0.082 453.591 0.966
26.78 3.83 0.024 144.704 0.961
35.074 5.681 0.02 123.93 0.962
16.077 0.518 -0.011 301.644 0.97
8.974 0.27 -0.068 415.614 0.967
26.274 0.953 0.002 224.658 0.97
8.044 0.204 -0.089 470.853 0.962
4.857 0.107 -0.186 643.636 0.959
7.983 0.194 -0.128 481.099 0.961
7.958 0.425 -0.029 352.213 0.965
11.41 0.275 -0.065 404.723 0.968
19.881 0.812 0.002 245.429 0.961
20.374 0.631 0.03 270.942 0.961
21.576 0.634 -0.024 269.724 0.966
24.525 1.079 0.022 215.473 0.969
21.821 0.594 0.005 277.163 0.967
175
13.92 0.422 -0.007 332.065 0.968
50.219 4.739 0.055 117.21 0.972
24.729 4.943 0.041 143.468 0.966
9.694 0.299 -0.028 396.245 0.969
23.853 2.55 0.075 164.251 0.97
68.252 7.269 0.094 97.515 0.969
16.39 0.617 -0.04 279.318 0.961
17.22 1.445 0 206.11 0.965
25.727 1.681 0.028 182.287 0.966
14.376 0.943 0.041 243.577 0.964
67.139 7.679 0.045 96.719 0.971
35.295 6.072 0.031 122.416 0.962
57.022 6.27 0.031 105.549 0.966
40.265 4.263 0.045 126.904 0.971
13.231 0.862 -0.023 253.779 0.958
7.922 0.219 -0.098 456.155 0.959
18.211 0.496 -0.015 303.827 0.968
21.102 1.042 -0.001 221.511 0.96
15.155 0.53 -0.009 298.833 0.96
12.269 0.347 -0.019 362.847 0.96
13.693 0.486 0.006 221.492 0.484
6.696 0.186 -0.115 495.577 0.961
29.495 3.63 0.007 142.928 0.97
49.366 6.035 0.015 110.972 0.972
48.964 7.208 0.013 107.329 0.971
34.71 2.844 0.002 146.545 0.971
37.805 2.239 0.016 155.889 0.972
17.773 0.644 -0.021 272.066 0.96
8.109 0.3 -0.113 401.496 0.965
18.046 2.007 -0.066 186.6 0.965
26.651 2.679 0.012 157.151 0.961
20.503 2.749 -0.028 167.425 0.96
15.404 1.484 0.048 209.306 0.963
6.733 0.152 -0.12 539.703 0.959
9.221 0.179 -0.086 494.978 0.965
13.749 0.594 -0.065 288.763 0.959
6.54 0.129 -0.182 583.633 0.963
65.269 7.173 0.022 98.922 0.969
Table B4: Indentation test of AAFA mix 4
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
45.581 11.046 -0.286 104.457 0.967
26.999 1.811 -0.35 176.182 0.963
176
28.713 4.548 -0.329 137.094 0.958
78.123 6.879 -0.327 95.358 0.959
23.754 1.331 -0.36 199.447 0.96
14.825 0.918 -0.397 244.674 0.965
73.081 9.023 -0.349 91.131 0.968
29.318 1.619 -0.354 181.119 0.968
44.656 2.281 -0.402 150.657 0.965
53.878 1.627 -0.431 168.876 0.971
34.691 4.063 -0.368 133.262 0.969
70.313 8.491 -0.311 93.313 0.968
90.064 10.165 -0.367 83.861 0.967
47.573 4.774 -0.408 118.039 0.965
35.711 3.326 -0.346 139.143 0.969
20.003 0.892 -0.27 237.563 0.969
58.44 6.528 -0.342 103.886 0.966
18.433 1.497 -0.403 200.718 0.962
17.747 0.473 -0.429 309.541 0.963
23.119 0.593 -0.359 274.817 0.957
41.298 3.206 -0.297 136.317 0.968
4.719 0.111 -0.502 636.555 0.967
25.981 1.641 -0.376 183.702 0.968
70.803 10.652 -0.322 89.13 0.969
9.86 0.142 -0.438 548.271 0.968
51.404 2.151 -0.332 150.877 0.96
30.255 3.913 -0.368 138.792 0.959
82.742 10.401 -0.334 85.594 0.972
22.297 1.963 -0.323 177.995 0.959
25.594 1.378 -0.357 195.405 0.966
31.821 5.766 -0.328 128.413 0.97
27.078 2.2 -0.303 165.394 0.96
52.125 3.231 -0.314 129.965 0.96
39.39 3.544 -0.206 133.881 0.97
23.469 1.213 -0.399 206.771 0.962
12.753 0.876 -0.406 254.797 0.964
29.205 1.532 -0.208 184.686 0.967
34.916 2.453 -0.393 152.331 0.96
13.15 0.522 -0.365 306.083 0.964
7.37 0.205 -0.454 473.039 0.962
17.57 0.623 -0.322 275.855 0.959
54.549 5.452 -0.313 110.602 0.968
36.844 1.975 -0.339 162.92 0.964
6.713 0.312 -0.269 404.474 0.969
32.965 1.647 -0.307 177.129 0.969
114.513 9.979 -0.26 79.584 0.965
12.949 0.429 -0.275 330.54 0.958
15.897 0.863 -0.348 246.693 0.959
22.697 1.238 -0.272 207.083 0.97
177
121.047 6.284 -0.136 91.353 0.968
36.441 1.99 -0.295 162.543 0.962
9.171 0.127 -0.301 579.451 0.967
45.041 4.132 -0.244 123.931 0.961
15.423 2.283 -0.261 190.325 0.97
11.736 0.279 -0.327 400.671 0.964
18.349 1.225 -0.239 214.634 0.967
97.121 13.002 -0.194 78.245 0.972
19.144 0.919 -0.215 235.88 0.967
61.083 8.117 -0.216 97.908 0.967
18.865 0.309 -0.214 372.011 0.963
31.72 1.934 -0.203 168.323 0.969
20.136 1.85 -0.168 185.921 0.968
46.508 2.197 -0.186 151.475 0.96
101.228 11.438 -0.184 79.17 0.965
106.348 7.263 -0.216 88.631 0.965
75.86 7.749 -0.152 93.551 0.97
32.425 2.011 -0.175 165.077 0.964
23.801 0.654 -0.207 264.512 0.968
62.665 4.972 -0.169 109.88 0.964
14.126 0.2 -0.245 461.558 0.969
7.876 0.15 -0.361 539.581 0.964
36.399 2.061 -0.131 160.462 0.961
31.675 1.666 -0.218 177.464 0.97
49.336 4.745 -0.107 117.471 0.969
30.082 1.131 -0.118 206.907 0.97
42.246 2.643 -0.077 144.698 0.971
24.222 1.359 -0.16 198.508 0.971
22.243 0.188 -0.14 465.73 0.96
28.172 1.951 -0.132 171.193 0.969
94.767 10.34 -0.094 82.617 0.97
28.107 1.548 -0.141 185.561 0.972
48.804 4.018 -0.112 123.496 0.971
59.773 3.505 -0.069 124.482 0.972
30.508 3.909 -0.091 139.334 0.97
74.778 10.879 -0.054 87.241 0.967
77.949 8.584 -0.063 90.652 0.97
21.566 1.683 -0.087 187.982 0.964
27.341 1.765 -0.087 176.957 0.96
43.538 2.654 -0.054 143.419 0.966
31.394 1.885 -0.089 170.162 0.97
57.706 8.329 -0.058 99.102 0.966
66.628 4.447 -0.068 112.779 0.97
29.435 0.905 -0.063 225.884 0.961
18.196 0.804 -0.114 249.906 0.968
27.091 0.505 -0.068 293.054 0.965
63.425 5.204 -0.063 108.512 0.971
178
13.715 0.189 -0.112 472.357 0.961
76.808 12.46 -0.024 84.495 0.965
59.308 6.328 -0.056 103.984 0.961
12.45 0.237 -0.081 430.34 0.971
35.395 1.76 -0.01 171.534 0.972
19.386 1.032 -0.021 225.427 0.964
38.186 2.972 -0.009 141.899 0.971
13.883 0.521 -0.066 305.347 0.97
33.037 2.06 -0.007 163.905 0.972
11.459 0.444 -0.07 330.838 0.963
47.48 3.792 0.008 125.7 0.962
19.572 0.687 -0.042 263.65 0.968
31.831 1.563 -0.012 181.436 0.969
78.247 7.592 0.021 93.214 0.967
16.72 0.277 -0.047 394.788 0.97
47.064 4.634 0.008 119.49 0.969
9.527 0.471 -0.087 331.753 0.97
71.448 7.474 0.014 95.945 0.973
14.384 0.46 -0.042 318.711 0.961
36.418 1.379 0.011 187.716 0.973
57.082 3.655 0.008 123.649 0.972
18.709 0.596 -0.017 280.89 0.971
15.98 0.503 -0.029 304.635 0.965
28.319 1.076 0.003 211.488 0.962
60.467 4.824 0.009 112.093 0.972
24.392 0.884 -0.055 232.335 0.963
26.449 1.513 -0.016 188.428 0.969
80.78 8.214 0.026 90.695 0.967
83.929 7.71 0.006 91.189 0.963
24.083 1.959 -0.029 175.255 0.959
16.71 0.677 -0.031 268.524 0.96
12.782 0.437 -0.06 330.325 0.969
27.26 1.695 -0.028 179.679 0.962
35.963 4.124 0.014 131.084 0.963
81.91 12.199 0.007 83.035 0.965
58.744 6.895 0.009 102.796 0.973
25.061 1.624 -0.018 184.888 0.962
19.664 1.166 -0.029 214.984 0.96
18.146 0.657 -0.041 269.351 0.96
13.411 0.326 -0.04 370.609 0.963
17.979 1.172 0.008 218.781 0.969
12.415 0.552 -0.048 300.591 0.959
13.315 0.738 -0.051 268.451 0.965
12.443 0.387 -0.084 347.949 0.968
25.944 3.008 -0.016 153.834 0.962
16.157 1.245 -0.035 218.847 0.97
18.766 0.616 -0.029 275.594 0.96
179
17.463 0.877 -0.024 241.951 0.959
85.062 10.533 0.019 84.537 0.967
17.145 0.905 -0.015 239.965 0.961
19.637 1.126 -0.041 217.632 0.96
12.495 0.423 -0.059 334.179 0.961
45.237 4.805 0.025 119.369 0.966
43.725 6.549 0.009 112.778 0.965
23.723 0.831 -0.038 238.844 0.962
17.277 0.837 -0.046 246.802 0.961
21.792 0.798 0.009 246.066 0.972
14.645 1.647 -0.035 205.98 0.959
20.361 0.869 -0.019 238.735 0.964
16.479 0.788 -0.043 254.649 0.966
27.281 0.605 -0.05 270.688 0.967
63.568 6.592 0.025 101.691 0.97
35.855 4.447 0.009 129.506 0.97
23.448 2.933 -0.006 159.122 0.963
84.396 9.003 0.025 87.862 0.97
19.947 0.857 -0.018 240.545 0.963
12.152 0.361 -0.057 358.973 0.97
24.143 1.454 -0.012 193.598 0.967
23.902 1.733 0.006 182.804 0.965
9.875 0.584 -0.069 303.762 0.958
50.861 6.945 0.018 106.581 0.967
10.588 0.189 -0.125 478.04 0.959
13.491 1.205 -0.032 228.018 0.959
19.308 0.681 -0.033 264.988 0.968
26.401 1.271 -0.025 199.915 0.961
13.4 0.579 -0.051 293.13 0.963
20.821 0.771 -0.027 250.091 0.968
28.375 1.185 -0.032 204.182 0.967
51.905 9.853 0.007 100.286 0.969
70.384 7.753 0.026 95.075 0.965
14.26 0.831 -0.039 254.791 0.965
45.175 4.349 0.02 122.661 0.969
19.115 0.243 -0.053 416.49 0.969
65.244 8.86 0.031 94.657 0.972
96.004 8.802 0.022 85.737 0.97
43.921 6.511 0.012 112.961 0.969
12.414 0.459 -0.03 324.905 0.97
15.447 1.593 -0.012 206.081 0.97
52.086 2.776 0.017 137.887 0.973
48.841 3.891 0.025 124.718 0.972
20.917 0.628 -0.036 272.02 0.969
29.439 2.767 0.02 153.05 0.971
86.907 8.901 0.04 87.593 0.972
37.378 2.24 -0.007 156.106 0.971
180
15.683 0.533 -0.047 297.771 0.963
14.281 0.438 -0.035 325.306 0.961
67.64 8.874 0.015 93.511 0.969
36.847 1.429 0.017 184.247 0.966
103.801 11.17 0.03 78.991 0.963
51.247 5.429 0.015 112.613 0.972
7.45 0.453 -0.077 347.336 0.962
20.431 1.095 -0.018 218.511 0.96
18.703 1.134 -0.015 219.932 0.97
Table B5: Indentation test of AAFA mix 5
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
14.579 0.606 0.343 284.978 0.962
10.087 0.298 0.326 393.545 0.961
57.892 6.69 0.403 103.9 0.973
85.726 10.808 0.355 83.965 0.969
53.235 5.796 0.341 109.277 0.963
21.758 0.603 0.268 276.129 0.971
55.164 5.143 0.324 112.093 0.97
53.488 2.692 0.27 138.242 0.963
28.426 0.982 0.231 220.097 0.969
23.655 2.956 0.209 158.249 0.961
19.374 0.781 0.153 249.753 0.96
26.336 1.562 0.205 186.765 0.972
26.641 1.145 0.19 208.525 0.969
29.673 1.638 0.182 179.697 0.964
18.698 2.235 0.142 180.12 0.964
21.75 0.303 0.112 373.811 0.966
24.499 1.113 0.149 212.959 0.968
55.394 2.372 0.152 144.667 0.968
31.843 1.868 0.106 170.344 0.97
35.019 2.65 0.118 148.692 0.961
55.06 6.226 0.116 106.877 0.97
96.376 7.757 0.101 88.434 0.963
28.118 1.582 0.064 183.465 0.965
36.894 1.949 0.07 163.892 0.967
20.63 0.799 0.055 245.769 0.959
19.291 0.459 0.03 313.119 0.971
32.609 1.904 0.031 168.573 0.97
51.026 5.312 0 79.236 0.48
17.029 0.55 -0.014 292.433 0.967
124.101 14.06 0.057 71.754 0.966
28.537 2.106 0.03 165.923 0.96
181
38.051 4.401 0.022 127.058 0.96
23.022 1.092 -0.027 216.389 0.971
46.247 3.719 -0.009 127.693 0.971
22.722 0.731 -0.011 252.908 0.964
22.335 0.701 -0.03 258.547 0.97
23.097 1.561 -0.023 190.23 0.965
25.749 1.535 -0.023 188.553 0.971
21.608 2.287 -0.015 172.926 0.969
10.877 0.396 -0.07 347.553 0.959
133.529 27.407 0.003 63.267 0.965
96.278 9.914 -0.029 83.123 0.97
96.669 8.056 -0.029 87.732 0.97
21.831 1.105 -0.068 215.838 0.96
13.447 0.481 -0.094 314.931 0.962
26.536 1.116 -0.068 210.226 0.964
48.607 2.564 -0.022 142.367 0.961
25.234 0.75 -0.015 248.263 0.967
16.735 0.292 -0.019 386.313 0.972
12.21 0.769 -0.097 267.829 0.963
18.249 0.678 -0.045 265.972 0.962
14.866 0.407 -0.068 335.516 0.967
26.572 2.234 0.008 165.743 0.965
95.169 8.407 0.01 86.981 0.97
17.48 0.806 -0.013 249.798 0.96
20.199 1.12 -0.036 217.453 0.963
10.511 0.309 -0.09 387.403 0.968
20.685 0.974 -0.043 228.712 0.968
21.042 0.929 -0.052 232.48 0.97
28.877 1.043 -0.011 214.389 0.968
21.297 1.089 -0.036 218.155 0.964
20.146 0.992 -0.051 226.996 0.961
96.903 9.045 0.014 84.841 0.968
17.291 0.64 -0.079 274.066 0.964
55.237 6.608 0.005 104.857 0.961
34.102 3.577 -0.005 137.52 0.962
19.349 1.365 -0.049 205.459 0.969
14.049 0.405 -0.09 338.042 0.969
19.693 0.836 -0.08 243.021 0.962
62.179 8.965 0.008 95.579 0.967
27.071 0.987 -0.034 220.549 0.968
9.802 0.12 -0.161 590.404 0.959
82.413 7.196 0.001 93.476 0.967
36.772 3.075 0.004 141.558 0.971
128.272 12.673 0.017 72.918 0.966
84.147 7.726 0.003 91.277 0.967
89.721 8.157 0.016 88.7 0.967
31.723 0.986 -0.045 217.404 0.968
182
92.466 11.596 0.001 80.766 0.963
30.042 0.686 -0.022 254.201 0.964
80.891 8.198 -0.008 90.953 0.972
69.504 7.594 -0.004 96.045 0.969
35.97 3.52 -0.037 136.718 0.968
113.712 14.549 -0.009 72.94 0.966
19.228 0.824 -0.056 246.098 0.97
22.232 0.692 -0.021 260.044 0.97
10.323 0.142 -0.14 545.42 0.961
10.655 0.253 -0.075 422.051 0.969
70.514 11.18 -0.003 88.592 0.97
81.125 9.056 -0.014 88.621 0.97
11.972 0.127 -0.067 574.299 0.97
39.357 4.393 -0.03 126.51 0.968
9.21 0.311 -0.159 389.668 0.96
22.847 2.102 -0.062 173.803 0.961
14.272 1.153 -0.098 229.542 0.97
69.102 7.588 -0.014 96.334 0.972
85.129 7.843 -0.035 90.675 0.967
43.109 3.767 -0.031 128.818 0.967
23.258 1.397 -0.04 197.566 0.968
99.29 9.157 -0.024 84.106 0.968
36.136 2.671 -0.074 147.663 0.964
24.082 0.897 -0.087 231.663 0.966
14.481 0.747 -0.078 263.334 0.959
33.206 2.709 -0.044 149.972 0.97
35.525 1.926 -0.034 165.459 0.967
14.642 0.345 -0.105 359.15 0.961
20.386 1.096 -0.086 218.519 0.96
56.05 10.461 -0.029 96.671 0.966
47.904 1.858 -0.035 161.536 0.967
38.756 5.824 -0.025 119.946 0.97
73.822 5.133 -0.014 105.809 0.97
36.089 1.392 -0.054 186.448 0.965
25.266 0.788 -0.019 243.648 0.97
33.472 2.633 -0.015 150.378 0.962
28.088 1.506 -0.053 187.125 0.969
21.017 1.469 -0.019 197.588 0.968
48.868 5.4 -0.019 113.431 0.96
67.904 8.529 -0.011 94.04 0.967
50.159 6.511 -0.018 108.395 0.967
22.992 1.61 -0.047 188.485 0.965
75.223 8.144 -0.01 92.61 0.97
78.143 11.19 0.003 85.752 0.97
25.519 1.262 -0.055 201.405 0.961
36.102 1.976 -0.054 163.814 0.97
21.801 0.323 -0.057 362.891 0.965
183
50.741 6.597 -0.016 108.067 0.972
17.045 1.02 -0.076 230.113 0.959
18.185 0.513 -0.076 298.344 0.961
8.553 0.434 -0.044 345.573 0.962
17.888 0.691 -0.082 264.722 0.963
114.719 9.852 -0.008 79.611 0.96
22.97 1.752 -0.039 183.472 0.964
71.994 12.477 -0.007 86.46 0.967
48.902 3.9 -0.015 124.526 0.971
29.584 0.761 -0.032 242.95 0.961
29.102 1.144 -0.028 206.564 0.969
21.604 1.034 -0.046 222.29 0.967
23.626 1.049 -0.066 218.677 0.968
49.935 2.648 -0.035 140.216 0.961
23.449 2.069 -0.022 174.424 0.97
13.934 0.263 -0.123 407.269 0.967
28.401 5.383 -0.027 134.645 0.964
23.005 1.088 -0.071 215.54 0.96
28.088 1.747 -0.037 177.71 0.97
28.766 1.195 -0.052 203.217 0.968
60.298 7.578 0.002 99.386 0.961
36.234 1.655 -0.053 174.737 0.97
10.987 0.475 -0.078 324.786 0.969
38.86 1.185 -0.035 197.897 0.968
11.273 0.459 -0.093 326.313 0.959
97.549 10.023 -0.007 82.534 0.968
102.87 11.512 -0.004 78.721 0.965
25.156 1.379 -0.032 196.259 0.969
33.795 1.079 -0.053 207.944 0.965
26.62 1.246 -0.052 202.011 0.969
13.439 0.645 -0.106 282.145 0.97
51.108 4.662 -0.003 117.084 0.969
11.382 0.688 -0.109 280.744 0.958
36.308 5.14 -0.007 125.135 0.968
30.023 5.017 -0.017 133.147 0.961
16.504 0.702 -0.055 264.978 0.959
14.3 0.776 -0.076 261.03 0.966
36.539 2.966 -0.02 142.543 0.962
40.176 2.313 -0.005 152.699 0.971
43.412 6.516 -0.007 113.456 0.971
98.169 10 -0.002 82.45 0.968
61.083 6.805 -0.017 101.853 0.969
36.106 1.644 -0.026 174.925 0.966
15.886 0.808 -0.081 252.665 0.96
14.672 0.541 -0.078 298.612 0.967
23.726 1.183 -0.042 208.762 0.966
21.302 1.088 -0.06 217.79 0.96
184
16.088 0.618 -0.079 280.01 0.963
59.069 4.544 -0.004 114.489 0.969
65.589 7.872 0.004 96.846 0.972
30.809 2.822 -0.003 149.785 0.96
37.805 4.733 -0.043 125.925 0.971
14.358 0.53 -0.093 300.855 0.962
11.893 0.667 -0.079 282.039 0.959
31.221 1.986 0.006 166.991 0.966
14.308 0.812 -0.067 256.161 0.961
8.103 0.176 -0.158 502.425 0.966
62.979 6.881 -0.002 100.384 0.961
14.577 0.189 -0.145 470.901 0.958
21.091 1.467 -0.036 197.818 0.971
29.932 1.631 -0.026 180.488 0.972
31.009 0.687 -0.013 253.139 0.961
31.469 1.943 -0.034 168.373 0.97
61.176 6.529 -0.012 102.964 0.971
46.364 3.695 -0.005 127.453 0.964
92.446 9.399 0.017 84.821 0.965
59.17 4.254 -0.003 116.955 0.972
98.681 10.133 0.005 82.088 0.968
32.904 3.135 0 143.806 0.965
38.726 1.295 -0.009 189.998 0.96
23.29 0.836 -0.054 238.532 0.961
38.694 3.283 -0.006 136.672 0.961
26.36 1.437 -0.029 192.335 0.972
28.801 2.239 -0.008 162.661 0.962
Table B6: Indentation test of AAFA mix 6
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
105.555 9.816 0.262 81.19 0.961
85.377 7.989 0.273 90.489 0.974
17.516 0.758 0.232 255.864 0.961
81.418 11.417 0.311 84.284 0.968
40.845 3.073 0.305 138.576 0.97
44.011 3.754 0.316 128.738 0.972
54.077 5.572 0.348 110.296 0.97
22.625 1.121 0.292 214.729 0.97
50.659 7.791 0.335 104.682 0.97
18.08 1.665 0.293 196.309 0.97
34.047 2.696 0.307 149.553 0.972
16.095 0.569 0.248 289.352 0.963
52.151 6.112 0.316 109.038 0.972
185
11.978 0.675 0.26 282.461 0.972
7.944 0.211 0.158 466.38 0.971
15.349 0.712 0.254 267.009 0.967
43.046 2.086 0.273 156.592 0.967
72.251 8.571 0.299 92.56 0.971
17.105 0.69 0.215 266.429 0.965
24.414 0.602 0.229 272.993 0.964
40.984 4.956 0.252 121.761 0.969
89.759 9.439 0.281 85.392 0.966
6.429 0.367 0.116 383.636 0.968
15.608 0.834 0.198 251.052 0.965
28.428 3.26 0.243 147.512 0.964
102.181 8.596 0.246 84.879 0.964
73.174 8.056 0.252 93.563 0.971
7.431 0.181 0.1 498.972 0.963
17.412 0.561 0.19 289.319 0.967
26.023 1.353 0.215 196.925 0.972
11.963 0.545 0.108 304.397 0.966
28.988 1.642 0.179 179.974 0.962
20.199 1.022 0.179 224.553 0.961
11.872 0.414 0.106 338.538 0.961
10.662 0.237 0.108 432.709 0.964
52.869 6.957 0.197 105.66 0.973
18.468 1.053 0.137 224.854 0.96
15.842 0.332 0.124 364.155 0.964
56.3 6.035 0.175 107.231 0.973
29.534 2.263 0.171 161.362 0.962
21.254 1.033 0.134 222.782 0.966
15.586 1.404 0.164 211.778 0.961
36.102 4.775 0.151 127.357 0.971
84.025 7.855 0.171 90.855 0.966
21.769 0.76 0.096 249.527 0.961
26.392 1.701 0.114 181.171 0.97
77.953 10.949 0.149 86.164 0.97
78.831 6.682 0.164 96.413 0.968
181.27 11.682 0.15 69.691 0.966
14.429 0.592 0.055 288.357 0.965
18.507 1.352 0.086 207.924 0.97
21.24 1.21 0.094 211.111 0.973
15.943 0.339 0.082 360.287 0.962
17.891 0.452 0.068 316.385 0.967
16.694 1.585 0.115 202.819 0.972
18.277 0.889 0.086 239.723 0.963
59.145 7.672 0.126 100.293 0.973
23.833 0.933 0.083 229.029 0.971
21.971 0.703 0.069 257.758 0.964
23.678 0.834 0.073 239.199 0.968
186
21.031 0.991 0.066 226.948 0.97
12.929 0.512 -0.019 309.283 0.967
19.064 0.94 0.056 234.587 0.971
23.771 1.354 0.082 198.781 0.966
21.429 0.979 0.05 227.367 0.969
13.88 0.374 -0.034 348.83 0.962
97.875 8.751 0.115 85.593 0.972
45.629 7.263 0.083 109.49 0.968
30.795 1.736 0.067 175.14 0.965
47.018 4.629 0.077 119.725 0.972
80.225 9.056 0.095 88.872 0.97
22.964 0.969 0.049 225.79 0.964
26.138 0.872 0.051 231.966 0.962
25.409 1.43 0.046 192.964 0.965
14.923 0.686 0.01 271.605 0.966
26.976 0.856 0.022 234.236 0.971
25.569 1.051 0.043 216.183 0.965
15.44 0.869 0.042 247.467 0.962
13.575 0.717 0.008 270.597 0.966
32.361 1.097 0.033 206.952 0.961
14.112 0.54 0.018 300.359 0.969
93.249 10.113 0.078 83.277 0.968
43.244 2.089 0.081 156.865 0.973
17.766 1.414 0.056 206.826 0.972
39.92 2.587 0.079 147.209 0.972
18.185 0.919 0.032 237.359 0.965
87.933 7.883 0.073 89.743 0.963
12.122 0.211 -0.029 454.041 0.969
37.476 3.999 0.056 130.875 0.966
21.044 0.933 0.019 231.412 0.963
15.78 0.532 0.021 298.294 0.966
84.003 9.638 0.066 86.602 0.97
40.784 4.546 0.056 124.362 0.969
69.713 7.453 0.06 96.08 0.963
24.818 3.243 0.052 153.392 0.965
19.847 1.416 0.053 201.359 0.962
15.712 0.462 -0.048 315.128 0.959
17.92 0.731 0 259.09 0.964
16.491 0.681 0.008 268.191 0.959
17.514 0.81 0.096 250.552 0.97
22.426 1.105 0.08 216.187 0.97
32.078 3.531 0.094 140.257 0.964
36.262 3.296 0.093 138.44 0.961
25.19 2.126 0.08 169.734 0.962
29.976 2.005 0.059 167.155 0.961
24.828 2.228 0.072 168.715 0.97
67.804 8.198 0.077 95.075 0.971
187
43.991 4.671 0.076 120.707 0.961
97.399 9.679 0.078 83.265 0.967
27.545 1.007 0.06 218.828 0.971
18.19 0.215 -0.037 442.547 0.971
37.047 4.222 0.041 129.703 0.968
32.124 4.056 0.06 135.651 0.962
20.445 0.567 0.003 283.645 0.964
14.396 0.933 0.007 244.593 0.966
16.241 0.856 -0.01 246.566 0.959
27.479 1.126 0.009 208.295 0.96
18.871 0.785 -0.02 250.764 0.966
86.427 9.775 0.055 85.557 0.967
19.355 0.99 0.004 228.733 0.963
66.341 10.544 0.066 91.082 0.967
63.017 8.255 0.072 96.696 0.967
15.934 0.55 -0.003 294.166 0.967
18.201 0.885 0 239.915 0.959
36.292 5.023 0.049 125.274 0.962
58.614 8.447 0.061 98.371 0.966
51.687 3.096 0.052 132.189 0.963
33.591 3.172 0.035 142.269 0.96
16.132 0.638 -0.008 276.243 0.962
81.443 8.019 0.045 91.184 0.969
27.309 1.38 0.003 194.173 0.972
12.184 0.379 -0.029 351.988 0.97
48.881 1.75 0.033 165.395 0.971
59.681 8.161 0.047 98.603 0.969
85.804 10.765 0.051 84.063 0.97
11.007 0.79 -0.005 270.086 0.961
30.86 1.599 0.026 179.835 0.96
20.342 1.047 -0.018 222.123 0.959
18.099 1.164 0.009 219.143 0.97
16.311 0.409 -0.008 331.922 0.964
18.17 0.394 -0.04 335.441 0.968
14.953 0.839 -0.015 252.113 0.965
123.694 12.015 0.037 74.414 0.963
32.95 1.087 0.018 207.309 0.96
238.395 21.722 0.029 55.453 0.963
80.081 10.955 0.048 85.567 0.973
20.005 1.613 0.028 193.088 0.962
77.969 7.712 0.04 92.719 0.963
68.571 9.797 0.047 91.276 0.967
5.428 0.094 -0.084 681.235 0.968
69.005 7.935 0.05 95.371 0.972
59.691 5.369 0.032 108.995 0.972
14.523 0.698 -0.017 270.852 0.966
50.334 3.301 0.031 130.661 0.973
188
25.963 1.597 0.004 185.48 0.969
34.533 2.198 0.014 158.689 0.966
55.062 6.872 0.017 104.055 0.961
27.184 1.907 0.001 173.516 0.968
20.086 0.587 -0.004 280.858 0.969
33.662 2.913 0.033 145.623 0.96
22.421 1.791 0.016 183.193 0.965
24.583 1.29 0.006 200.628 0.96
27.026 1.274 -0.003 200.22 0.972
16.804 0.972 -0.036 234.713 0.961
23.681 1.155 -0.006 210.019 0.96
31.103 0.844 0.003 231.83 0.963
73.229 8.255 0.031 92.942 0.969
4.692 0.289 0.026 438.551 0.972
19.138 1.372 0.039 204.91 0.963
32.707 5.083 0.024 129.303 0.964
18.44 0.964 -0.012 233.387 0.97
47.695 6.366 0.029 110.426 0.965
24.755 1.246 -0.012 203.131 0.961
17.357 1.116 -0.028 222.73 0.961
91.537 6.035 0.028 96.933 0.973
94.679 8.969 0.037 85.583 0.97
51.28 5.71 0.032 111.234 0.972
13.592 0.444 -0.057 325.211 0.963
15.997 0.497 -0.019 306.439 0.966
21.856 1.223 -0.011 208.947 0.969
48.954 2.858 0.026 137.649 0.971
58.675 6.645 0.032 103.557 0.97
9.703 0.208 -0.007 461.978 0.968
33.399 2.289 0.041 157.911 0.971
17.86 0.933 -0.008 236.033 0.961
59.077 5.66 0.031 107.578 0.97
20.74 1.026 -0.028 223.197 0.959
8.726 0.264 -0.074 418.974 0.96
98.447 7.685 0.043 88.479 0.967
10.405 0.196 -0.095 469.513 0.957
12.547 0.263 -0.078 409.72 0.964
22.024 0.915 -0.009 232.822 0.972
13.469 0.728 0.004 268.521 0.96
20.295 0.802 -0.004 247.144 0.969
9.985 0.455 -0.043 333.414 0.965
16.535 0.75 -0.026 258.945 0.963
25.855 3.585 0.017 148.211 0.961
17.937 0.675 -0.024 267.304 0.964
12.619 0.667 -0.051 279.937 0.962
189
Table B7: Indentation test of AAFA mix 7
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
23.339 1.363 0.041 199.534 0.972
13.466 0.356 -0.052 356.684 0.961
68.245 10.411 0.048 90.461 0.967
17.314 0.87 -0.042 243.406 0.962
16.26 0.646 -0.064 275.262 0.966
31.659 4.191 0.001 135.131 0.96
10.203 0.488 -0.05 323.111 0.962
79.47 8.419 0.026 90.505 0.967
16.133 0.712 -0.043 265.868 0.971
20.831 1.154 -0.008 215.192 0.972
16.833 0.828 -0.04 248.761 0.962
49.675 5.141 0.022 114.876 0.969
36.088 5.472 0.014 123.524 0.962
12.479 0.577 -0.07 297.379 0.971
11.987 0.395 -0.104 345.633 0.965
7.637 0.426 -0.05 354.149 0.966
17.55 0.698 -0.054 264.746 0.966
16.491 0.61 -0.034 281.105 0.966
51.726 4.987 0.018 114.672 0.969
17.439 1.077 -0.051 226.439 0.971
66.858 7.989 0.022 95.969 0.97
11.115 0.454 -0.06 330.123 0.97
15.082 0.65 -0.058 276.952 0.966
17.085 0.72 -0.049 261.809 0.962
115.678 10.744 0.024 77.756 0.963
80.133 7.089 0.025 94.505 0.968
17.697 0.531 -0.074 295.268 0.965
16.308 0.9 -0.045 242.558 0.963
21.305 0.748 -0.031 251.859 0.963
18.54 0.588 -0.019 281.094 0.959
16.54 1.146 -0.029 222.968 0.964
30.584 2.831 -0.008 150.214 0.964
40.104 5.434 0.001 119.746 0.962
67.964 9.934 0.025 91.431 0.97
46.986 3.908 0.018 125.531 0.972
29.161 0.626 -0.027 264.651 0.961
38.512 3.545 0.008 133.887 0.961
107.601 11.42 0.028 77.883 0.963
15.191 0.793 -0.056 256.522 0.964
20.262 0.881 -0.042 237.142 0.96
14.078 0.715 -0.052 269.052 0.963
73.392 9.025 0.031 90.955 0.967
18.04 0.869 -0.035 243.113 0.97
190
16.849 0.788 -0.053 253.292 0.961
14.12 0.647 -0.051 279.219 0.963
13.173 0.611 -0.068 288.142 0.965
22.132 1.599 -0.021 190.55 0.969
75.509 9.188 0.023 89.911 0.967
21.507 1.035 -0.022 221.912 0.964
22.151 0.654 -0.053 265.127 0.962
9.465 0.117 0 598.35 0.963
42.163 3.195 0.071 136.191 0.971
24.601 0.676 -0.006 259.035 0.96
14.583 0.636 -0.046 279.876 0.963
15.453 0.683 -0.044 271.455 0.97
16.045 0.764 -0.04 258.157 0.964
63.08 3.937 0.048 118.524 0.97
12.398 0.692 -0.055 278.214 0.97
61.963 6.208 0.012 104.014 0.973
22.115 0.867 -0.023 236.4 0.961
74.116 10.166 0.023 88.679 0.97
16.027 0.832 -0.055 249.984 0.962
28.037 0.965 0.024 222.302 0.972
17.812 0.491 -0.038 305.01 0.965
16.213 0.664 -0.048 272.034 0.964
17.786 0.791 -0.048 250.84 0.959
19.017 0.921 -0.034 235.071 0.96
15.673 0.739 -0.052 262.084 0.963
12.745 0.663 -0.045 280.25 0.962
43.255 3.779 0.016 128.178 0.96
65.052 7.43 0.021 98.313 0.972
6.296 0.338 -0.123 396.316 0.969
14.694 0.691 -0.046 270.851 0.963
81.251 11.757 0.029 83.884 0.968
16.024 0.753 -0.055 259.544 0.964
58.129 9.122 0.026 97.346 0.967
20.826 0.773 -0.008 249.345 0.964
56.645 3.914 0.035 121.098 0.973
44.758 5.231 0.019 117.466 0.969
24.692 0.886 -0.004 231.621 0.961
78.219 13.682 0.034 82.977 0.968
16.037 0.406 -0.088 333.061 0.962
19.06 0.838 -0.039 243.485 0.959
38.09 7.143 0.018 116.851 0.968
30.269 2.259 0.029 161.452 0.971
9.24 0.25 -0.038 426.619 0.96
58.586 7.252 0.013 101.536 0.969
14.252 0.339 -0.044 362.301 0.958
39.215 2.049 0.01 160.023 0.972
24.949 1.656 0.021 184.624 0.97
191
10.669 0.215 -0.121 450.004 0.957
12.656 0.633 -0.054 285.81 0.965
6.504 0.196 -0.065 487.996 0.966
15.819 0.818 -0.043 252.536 0.967
16.325 0.653 -0.051 274.168 0.968
14.876 0.835 -0.032 251.867 0.959
18.907 0.7 -0.04 261.348 0.959
12.796 0.351 -0.046 360.354 0.962
14.271 1.007 0.02 238.701 0.964
70.532 14.388 -0.006 85.556 0.965
45.148 3.176 0.001 134.081 0.962
24.349 0.574 -0.055 278.497 0.963
23.342 1.145 -0.058 211.115 0.96
18.765 0.86 -0.06 241.527 0.959
69.832 8.318 -0.001 94.078 0.972
28.485 2.03 -0.011 168.45 0.966
17.543 0.822 -0.054 247.992 0.961
39.872 3.44 0.007 134.019 0.961
17.186 0.635 -0.056 274.894 0.962
26.746 0.73 -0.03 250.078 0.968
23.244 1.578 -0.04 189.966 0.971
11.157 0.53 -0.079 311.127 0.97
18.804 0.767 -0.055 252.636 0.962
82.892 6.222 0.007 97.128 0.961
114.49 11.891 0.017 76.245 0.97
36.797 1.17 -0.025 199.182 0.961
12.413 0.423 -0.08 335.512 0.969
26.21 1.275 -0.035 200.696 0.969
15.527 0.742 -0.02 262.255 0.965
32.229 1.551 -0.012 181.795 0.97
39.776 1.189 -0.036 197.091 0.967
45.752 1.334 -0.012 185.593 0.966
37.885 3.469 0.002 135.917 0.971
36.956 1.098 0.007 205.075 0.967
29.114 4.284 -0.006 138.091 0.961
26.154 0.889 -0.016 231.359 0.972
67.431 7.051 -0.005 98.459 0.968
25.743 1.655 -0.014 183.781 0.972
17.446 0.789 -0.045 252.383 0.964
16.755 0.68 -0.053 269.335 0.97
28.624 0.584 -0.064 274.118 0.967
12.921 0.964 -0.056 246.71 0.965
23.943 1.516 -0.017 191.642 0.972
84.873 8.247 0.002 89.554 0.967
32.802 0.677 -0.043 254.934 0.969
67.128 9.055 0.012 93.473 0.972
18.648 0.79 -0.044 249.9 0.961
192
72.54 9.509 0.011 90.399 0.97
27.184 1.505 -0.003 187.4 0.962
105.707 11.759 0.019 77.697 0.963
37.827 1.508 -0.011 180.156 0.969
116.946 11.177 0.015 76.763 0.963
77.226 9.439 0.022 88.958 0.97
25.032 1.016 -0.024 220.072 0.969
93.498 8.437 0.013 87.108 0.967
32.2 2.197 -0.003 161.032 0.971
39.489 4.394 0.017 125.881 0.96
17.989 0.737 -0.058 258.014 0.963
48.283 5.58 0.035 113.047 0.963
19.842 0.963 -0.047 229.951 0.96
17.159 0.698 -0.058 265.456 0.966
11.314 0.484 -0.099 319.666 0.959
41.523 7.239 0.013 112.636 0.96
16.245 0.481 -0.064 309.193 0.961
44.35 6.18 0.014 113.879 0.972
19.229 0.716 -0.052 258.619 0.959
17.02 0.773 -0.046 254.827 0.962
20.347 1.044 -0.058 222.393 0.959
23.301 1.283 -0.027 203.414 0.968
28.825 2.694 -0.004 154.262 0.963
29.028 0.733 -0.023 247.247 0.961
14.62 1.102 -0.029 230.633 0.961
42.471 5.135 0.005 119.39 0.965
53 6.633 0.022 106.526 0.971
24.874 0.908 -0.032 229.264 0.961
61.385 7.873 0.028 98.493 0.969
19.705 0.798 -0.056 247.237 0.96
17.74 0.589 -0.059 282.698 0.964
29.863 1.386 -0.022 190.587 0.962
22.442 1.148 -0.037 213.212 0.971
22.974 1.293 -0.038 203.355 0.968
26.657 0.842 -0.05 234.983 0.962
50.777 7.812 0.022 104.041 0.961
29.393 1.982 -0.013 169.023 0.969
42.764 1.622 0.007 172.43 0.966
29.316 1.162 -0.033 204.249 0.961
17.629 0.849 -0.039 244.57 0.959
72.477 13.427 0.027 85.45 0.967
93.379 9.069 0.027 85.495 0.967
63.297 3.753 -0.017 119.8 0.961
66.74 8.148 0.029 95.67 0.972
15.705 0.762 -0.074 259.44 0.966
32.146 1.164 0.007 202.472 0.964
17.171 0.787 -0.038 254.105 0.971
193
18.199 0.787 -0.061 251.62 0.966
21.676 0.752 -0.032 250.884 0.962
20.846 0.792 -0.045 246.649 0.962
21.654 1.224 -0.038 209.431 0.971
16.987 0.836 -0.039 248.14 0.967
57.357 6.652 0.014 104.136 0.97
78.703 9.017 0.003 89.4 0.97
74.115 10.154 0.013 88.701 0.97
26.636 2.368 -0.012 162.444 0.96
14.63 0.601 -0.072 287.232 0.971
17.436 0.692 -0.035 265.008 0.96
17.488 0.848 -0.045 245.231 0.962
19.882 1.023 -0.054 225.964 0.97
15.38 0.703 -0.068 268.699 0.97
75.48 8.407 0.009 91.705 0.967
Table B8: Indentation test of AAFA mix 8
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
2.604 0.023 -0.871 1326.701 0.945
2.756 0.057 -0.409 876.466 0.949
6.352 0.137 -0.196 569.45 0.965
26.454 1.398 -0.001 193.682 0.967
16.865 1.076 -0.015 226.385 0.959
89.099 7.778 0.046 89.807 0.963
24.599 2.133 -0.015 170.23 0.96
37.871 1.88 0.043 165.898 0.972
103.24 15.213 0.048 74.439 0.966
17.424 0.659 -0.028 270.587 0.963
26.779 1.593 -0.012 185.037 0.972
74.091 8.301 0.042 92.428 0.967
18.63 0.832 -0.036 245.092 0.962
69.244 7.317 0.049 96.676 0.963
15.035 0.708 -0.042 268.795 0.971
73.703 6.096 0.046 100.479 0.97
18.179 1.039 -0.027 226.429 0.96
11.979 0.733 -0.052 273.123 0.963
7.071 0.652 -0.159 312.616 0.959
2.274 0.024 -1.32 1306.035 0.943
4.139 0.105 -0.285 657.908 0.965
10.223 0.651 -0.085 293.062 0.971
23.223 1.268 -0.014 204.786 0.972
21.527 0.751 -0.028 250.993 0.96
79.429 9.391 0.02 88.266 0.967
194
16.373 0.707 -0.052 265.406 0.966
91.29 15.333 0.032 77.278 0.963
63.479 13.8 0.022 89.802 0.971
24.474 1.98 0.005 175.081 0.97
13.191 0.166 -0.064 502.014 0.962
61.669 8.464 0.041 97.098 0.972
24.476 1.261 -0.038 203.528 0.97
85.647 10.219 0.025 85.156 0.973
16.594 0.839 -0.059 248.133 0.963
53.855 3.433 0.028 127.105 0.967
14.667 0.654 -0.052 277.123 0.966
15.086 0.292 -0.043 386.026 0.962
14.924 0.657 -0.05 276.106 0.966
26.892 3.013 0.002 152.459 0.964
34.7 4.843 -0.001 128.042 0.964
16.72 1.161 0 221.157 0.96
19.841 1.763 0.017 189.009 0.967
26.469 0.762 -0.05 244.746 0.959
30.545 0.913 -0.016 225.418 0.97
14.596 0.408 -0.072 336.091 0.97
85.858 8.638 0.016 88.27 0.967
5.165 0.029 -0.723 1175.529 0.96
12.649 0.569 -0.063 296.456 0.959
48.64 6.626 0.022 108.648 0.961
25.201 0.93 -0.033 226.66 0.96
24.834 1.172 0.008 208.643 0.97
111.63 11.291 0.028 77.826 0.973
24.337 1.691 -0.033 183.277 0.961
14.241 0.702 -0.048 271.227 0.968
37.714 1.546 0.006 178.627 0.971
19.209 0.843 -0.003 244.142 0.971
34.46 1.441 0.02 185.002 0.966
13.494 0.463 -0.059 319.301 0.959
9.687 0.261 -0.083 418.504 0.966
28.825 3.223 0.016 147.47 0.965
19.257 0.947 -0.053 233.414 0.969
15.73 0.905 -0.041 243.38 0.963
82.581 9.417 0.036 87.314 0.967
51.106 5.589 0.033 111.908 0.972
37.125 2.308 0.01 153.866 0.961
39.431 2.658 0.01 145.298 0.96
82.425 7.445 0.023 92.746 0.97
84.813 7.729 0.041 91.225 0.969
15.898 0.711 -0.032 265.772 0.966
11.28 0.438 -0.077 332.79 0.96
12.837 1.265 -0.041 228.167 0.964
4.561 0.037 0.031 1051.81 0.971
195
56.65 3.518 0.034 124.634 0.961
84.451 10.084 0.041 85.497 0.968
18.626 0.55 -0.038 289.503 0.964
14.269 0.567 -0.023 294.183 0.969
49.84 7.073 0.046 107.089 0.971
12.711 0.709 -0.033 273.681 0.962
5.589 0.033 -0.266 1109.641 0.955
14.787 0.598 -0.023 286.049 0.963
15.99 0.916 0.012 241.259 0.959
63.778 2.062 0.028 105.937 0.483
52.121 5.626 0.068 110.938 0.968
14.2 0.674 -0.02 274.787 0.963
27.502 1.396 0.055 192.851 0.969
12.394 0.622 -0.021 288.483 0.965
16.715 0.661 0.018 271.61 0.964
23.76 0.457 -0.006 309.755 0.971
4.778 0.251 -0.081 457.112 0.961
17.469 0.984 0.019 232.586 0.963
17.857 1.064 0 225.326 0.961
13.018 0.509 -0.004 310.498 0.971
16.881 0.796 0.009 252.595 0.964
34.077 3.416 0.072 139.367 0.965
5.752 0.086 -0.135 703.76 0.965
35.205 4.845 0.057 128.023 0.972
16.903 0.799 0.011 252.569 0.966
14.397 0.576 -0.003 291.645 0.966
33.463 1.372 0.014 189.808 0.973
12.463 0.422 -0.038 333.992 0.959
17.651 0.816 0.029 248.87 0.964
13.238 0.618 -0.043 286.141 0.962
20.58 2.555 0.031 169.842 0.96
19.577 0.617 -0.015 274.503 0.961
32.847 1.96 0.039 165.855 0.96
9.969 1.004 0.023 258.121 0.969
14.873 0.647 -0.031 277.248 0.962
74.408 8.374 0.07 92.19 0.968
22.606 1.11 0.017 215.475 0.97
103.923 16.877 0.07 72.905 0.961
14.131 0.68 -0.064 275.039 0.97
12.346 0.713 -0.043 275.507 0.97
19.86 0.802 -0.028 247.068 0.964
16.043 0.827 -0.018 251.181 0.967
11.175 0.454 -0.008 329.702 0.967
19.223 1.186 0.008 214.497 0.96
12.913 0.756 -0.017 267.77 0.968
69.478 8.943 0.06 92.535 0.967
9.739 0.333 -0.066 376.6 0.959
196
12.303 1.334 -0.019 228.063 0.97
52.037 6.981 0.043 105.915 0.97
15.579 1.059 -0.022 231.93 0.969
55.247 3.887 0.028 121.652 0.969
40.627 6.712 0.023 115.059 0.965
14.453 0.684 -0.044 273.759 0.972
5.755 0.296 -0.146 419.667 0.963
19.361 0.87 0.001 240.232 0.965
98.481 11.488 0.048 79.559 0.963
16.993 0.851 -0.022 245.544 0.959
34.052 2.061 0.008 162.093 0.96
15.233 0.362 -0.046 350.654 0.961
21.662 1.384 -0.002 199.777 0.962
13.72 0.67 -0.052 276.485 0.962
12.475 0.507 -0.053 310.87 0.961
60.787 7.587 0.037 99.668 0.971
12.601 0.478 -0.035 317.642 0.961
28.457 0.766 0.012 244.234 0.971
25.952 1.039 -0.014 216.423 0.962
15.677 1.112 0 226.772 0.959
82.171 12.044 0.04 83.359 0.97
69.979 10.43 0.059 89.71 0.967
44.78 8.732 0.049 107.01 0.961
36.906 2.176 0.039 158.104 0.972
15.297 0.842 -0.014 250.814 0.964
66.575 6.947 0.041 99.261 0.97
10.15 0.327 -0.03 378.58 0.963
20.354 0.918 0.041 233.335 0.96
101.598 14.55 0.239 75.483 0.968
77.807 6.98 -0.024 95.575 0.969
71.466 8.847 0.055 92.015 0.967
31.153 5.592 0.052 129.934 0.97
50.068 6.165 0.046 109.875 0.97
15.895 0.597 -0.029 284.123 0.963
30.848 3.751 0.037 140.323 0.972
20.56 1.235 0.032 209.769 0.965
88.049 9.103 0.057 86.609 0.967
30.722 1.276 0.042 196.818 0.969
79.4 7.756 0.068 92.083 0.961
14.207 0.567 -0.015 294.587 0.97
18.319 0.876 -0.023 241.125 0.964
89.129 11.561 0.069 81.884 0.968
54.254 9.388 0.053 98.932 0.961
24.439 0.718 0 253.999 0.97
11.681 0.575 -0.04 299.179 0.964
12.186 0.47 -0.041 321.638 0.965
25.836 0.819 0.018 238.496 0.964
197
52.156 5.346 0.064 112.57 0.972
79.816 8.288 0.073 90.758 0.967
97.352 9.56 0.136 83.603 0.969
48.244 4.493 0.103 119.955 0.971
67.995 8.592 0.083 93.876 0.967
16.674 0.581 -0.021 286.312 0.966
113.953 12.962 0.082 74.472 0.96
21.22 1.009 0.058 224.3 0.962
76.426 9.768 0.056 88.466 0.967
12.398 0.543 -0.037 302.971 0.962
22.862 1.406 0.056 197.253 0.964
37.925 1.134 0.037 201.711 0.965
12.588 1.376 0.012 158.049 0.482
12.6 0.635 -0.051 284.975 0.961
58.504 7.027 0.076 102.334 0.97
11.176 0.619 -0.024 293.833 0.969
97.688 8.465 0.072 86.395 0.971
12.681 0.548 -0.019 302.334 0.969
28.284 2.997 0.051 150.524 0.961
11.1 0.444 -0.012 333.198 0.97
26.711 1.681 0.039 181.151 0.965
35.328 2.615 0 148.961 0.96
23.845 0.879 0.007 233.336 0.962
9.537 0.425 -0.06 343.65 0.964
26.842 0.961 -0.01 222.198 0.96
14.695 0.807 -0.029 255.961 0.963
21.12 1.183 0.012 211.613 0.961
13.909 0.673 -0.076 194.352 0.482
67.108 4.325 0.006 79.541 0.48
65.696 9.293 0.037 93.374 0.967
75.214 7.043 0.006 95.99 0.968
14.248 0.628 -0.064 282.649 0.967
0.685 0.005 -0.392 2844.819 0.961
Table B9: Indentation test of AAFA mix 9
Modulus Hardness Drift Correction Displacement Load
GPa GPa nm/s nm mN
17.648 0.57 -0.047 286.114 0.96
20.956 1.056 -0.044 221.753 0.97
42.156 3.119 0.02 136.718 0.964
29.446 3.293 0.014 145.56 0.961
46.601 6.578 0.011 110.785 0.971
13.15 0.382 -0.088 346.625 0.96
17.955 0.893 -0.04 239.531 0.959
198
19.887 1.019 -0.031 225.466 0.963
39.806 3.596 0.006 133.109 0.971
84.331 9.248 0.026 87.259 0.968
22.026 1.308 -0.029 203.158 0.962
37.528 1.471 0.006 181.762 0.966
65.714 7.887 0.01 96.742 0.972
75.647 5.455 0.025 103.487 0.973
19.996 1.001 -0.036 226.896 0.964
17.728 0.898 -0.053 239.463 0.959
29.402 1.806 -0.01 173.838 0.963
21.2 0.849 -0.027 239.217 0.959
53.518 3.666 0.024 124.877 0.972
23.913 1.129 -0.014 212.358 0.968
17.793 0.559 -0.022 289.682 0.97
28.66 1.062 -0.013 213.436 0.972
28.551 1.251 -0.025 199.143 0.961
26.932 1.126 -0.028 209.867 0.971
58.538 9.788 0.016 96.057 0.966
40.294 1.915 -0.023 163.153 0.969
90.852 14.454 0.028 78.276 0.968
6.71 0.117 -0.256 610.431 0.968
61.285 7.455 0.019 99.748 0.97
16.944 0.836 -0.044 248.712 0.97
26.673 0.945 -0.026 223.904 0.961
8.891 0.036 -0.728 1054.105 0.95
22.457 2.692 0.027 163.96 0.961
14.719 0.977 -0.062 239.354 0.961
15.922 0.758 -0.066 260.109 0.971
26.246 1.174 -0.024 207.179 0.971
20.566 0.694 -0.039 260.462 0.962
12.823 0.42 -0.101 334.851 0.965
42.963 2.621 0.001 144.671 0.97
80.8 13.263 0.026 82.409 0.967
16.766 0.794 -0.038 253.311 0.966
36.002 2.514 -0.001 151.197 0.971
13.479 0.708 -0.048 271.493 0.962
26.271 1.704 -0.017 181.264 0.97
16.311 0.543 -0.054 294.194 0.961
16.211 0.545 -0.045 294.981 0.968
16.88 0.7 -0.037 264.732 0.96
27.021 1.387 -0.021 193.033 0.962
20.55 0.941 -0.035 230.859 0.96
68.348 7.537 0.026 96.746 0.972
24.389 2.014 0.023 173.608 0.962
40.906 0.955 -0.03 215.317 0.961
16.144 0.797 -0.045 254.067 0.965
17.912 0.906 -0.051 239.133 0.966
199
26.976 2.604 0.005 158.441 0.968
19.21 1.194 -0.029 215.134 0.97
18.249 0.868 -0.041 242.161 0.964
33.483 5.676 0.037 126.22 0.967
19.329 0.942 -0.047 233.091 0.964
21.667 0.906 -0.04 232.768 0.96
17.433 0.913 -0.04 238.482 0.959
20.168 0.99 -0.035 227.065 0.96
17.553 0.886 -0.043 240.872 0.959
34.649 2.085 0.003 161.628 0.968
18.013 1.007 -0.04 229.754 0.964
15.729 0.724 -0.068 264.251 0.965
22.781 1.393 -0.015 197.674 0.961
21.672 1.38 -0.013 199.898 0.961
11.155 0.456 -0.057 329.281 0.969
22.944 0.757 -0.024 249.858 0.97
22.111 1.305 -0.036 203.068 0.961
33.411 3.849 0.005 136.537 0.973
24.002 0.72 -0.037 253.068 0.962
19.309 0.865 -0.045 240.597 0.963
89.628 10.684 0.027 83.085 0.967
13.112 0.93 -0.037 248.407 0.962
13.93 0.466 -0.061 317.87 0.962
16.073 0.588 -0.064 285.904 0.966
7.525 0.07 -0.146 771.243 0.969
25.741 3.089 -0.002 152.983 0.959
20.362 0.888 -0.007 236.433 0.961
28.065 3.781 0.009 143.115 0.962
143.541 13.554 0.039 69.79 0.963
17.923 0.933 -0.047 236.058 0.962
15.506 0.525 -0.055 299.267 0.959
16.287 1.247 -0.011 218.078 0.967
15.184 0.716 -0.053 267.24 0.97
18.443 0.851 -0.038 243.635 0.964
18.835 0.954 -0.033 232.754 0.964
27.607 1.428 0.002 191.119 0.968
18.948 0.837 -0.057 244.701 0.967
52.906 8.584 0.035 101.647 0.971
21.576 1.251 -0.023 207.973 0.971
17.268 0.977 -0.044 233.715 0.964
31.322 1.847 -0.003 170.92 0.964
16.055 0.992 -0.028 235.977 0.971
26.763 0.976 -0.041 222.206 0.971
48.662 1.88 0.015 160.931 0.972
19.026 0.907 -0.03 236.349 0.959
17.686 0.964 -0.031 234.237 0.966
57.09 1.3 0.057 184.989 0.97
200
34.167 1.449 0.032 185.077 0.969
44.707 6.65 0.038 111.391 0.96
82.219 10.139 0.04 86.014 0.967
31.959 6.39 0.012 126.241 0.964
17.794 0.751 -0.036 256.361 0.962
18.123 0.857 -0.037 244.199 0.97
13.904 0.585 -0.039 291.991 0.972
74.684 9.264 0.03 90.136 0.97
13.064 0.591 -0.062 291.396 0.961
20.871 0.659 -0.013 265.593 0.961
19.332 1.135 -0.012 218.784 0.971
26.674 0.583 -0.019 274.972 0.964
54.924 5.218 0.031 111.921 0.972
51.741 4.168 0.028 120.557 0.969
10.286 0.429 -0.077 338.93 0.96
105.4 11.871 0.052 77.705 0.966
33.057 5.768 0.026 126.664 0.969
55.671 6.95 0.036 104.14 0.973
29.858 4.793 0.021 134.356 0.961
18.528 0.742 -0.028 256.98 0.966
91.605 10.26 0.034 83.364 0.968
16.097 1.204 -0.003 220.875 0.966
196.967 14.625 0.039 64.36 0.972
12.307 1.003 -0.029 245.927 0.964
20.69 0.35 -0.047 351.357 0.97
53.807 6.331 0.025 107.262 0.972
16.735 1.846 0.02 193.826 0.962
43.188 5.371 0.019 118.086 0.972
25.141 0.871 -0.016 233.08 0.963
76.295 12.222 0.035 85.159 0.97
30.397 1.453 -0.013 187.661 0.97
28.849 1.128 -0.006 207.499 0.965
26.32 4.899 0.026 140.308 0.966
41.85 9.217 0.025 109.867 0.971
69.5 5.517 0.03 104.679 0.97
76.379 9.685 0.036 88.759 0.97
47.133 5.489 0.02 114.751 0.972
36.948 2.448 0.026 151.011 0.961
77.601 11.108 0.024 86.045 0.97
19.206 1.012 -0.025 227.899 0.97
19.042 0.896 -0.029 237.915 0.964
18.562 0.882 -0.034 240.095 0.964
66.397 4.294 0.015 114.12 0.969
55.887 4.035 0.021 119.561 0.962
16.168 1.204 -0.048 221.125 0.97
13.794 0.827 -0.028 255.795 0.96
21.094 1.07 -0.035 219.327 0.96
201
20.16 0.91 -0.038 234.232 0.959
10.63 0.229 -0.024 439.763 0.966
16.705 0.767 -0.035 255.895 0.96
39.576 2.28 0.029 153.001 0.961
97.637 12.504 0.033 78.662 0.971
18.017 0.896 0.026 239.61 0.964
21.798 1.105 -0.025 217.052 0.971
40.88 4.646 0.003 123.797 0.971
30.742 0.949 -0.016 221.583 0.969
34.363 1.662 -0.012 175.656 0.97
44.575 2.703 0.019 142.346 0.971
60.003 4.328 0.033 115.468 0.962
18.637 0.978 -0.041 231.091 0.964
33.539 2.625 0.025 150.574 0.964
22.44 1.403 -0.017 197.879 0.963
24.868 1.289 -0.013 200.511 0.961
35.886 3.836 0.014 134.005 0.97
5.335 0.075 -0.174 753.471 0.961
19.621 1.033 -0.035 224.405 0.96
55.249 4.258 0.026 118.34 0.97
17.525 0.559 -0.018 288.472 0.96
17.545 0.986 -0.024 232.251 0.963
59.129 9.409 0.031 96.457 0.969
31.041 1.751 -0.002 174.378 0.965
19.559 0.995 -0.028 228.754 0.97
49.454 3.669 0.026 126.655 0.972
35.843 4.844 0.031 127.248 0.971
25.944 2.573 0.028 160.444 0.968
15.952 0.677 -0.042 270.433 0.964
20.022 0.955 -0.02 230.451 0.961
20.374 0.885 -0.002 236.535 0.96
21.303 0.872 -0.014 236.766 0.96
31.539 3.994 0.023 136.819 0.962
23.069 0.619 -0.031 271.415 0.968
57.987 8.356 0.034 99.163 0.972
60.215 5.575 0.02 107.704 0.972
26.605 1.672 0.007 181.974 0.969
32.72 2.619 0 152.045 0.972
100.943 10.151 0.048 81.6 0.968
28.892 2.225 0.032 163.637 0.971
19.538 1.061 -0.028 222.406 0.96
24.361 1.381 -0.009 197.06 0.969
29.878 2.54 0.017 155.949 0.967
61.818 6.565 0.042 102.498 0.97
26.042 1.018 -0.008 218.901 0.969
7.875 0.06 -0.209 825.304 0.964
133.226 12.14 0.044 72.994 0.963
202
88.916 13.686 0.046 79.502 0.968
15.226 0.718 -0.032 265.82 0.963
24.186 3.809 0.036 149.575 0.96
77.749 7.224 0.051 94.521 0.966
203
References
Alonso, C. and Fernandez, L. (2004), ‘Dehydration and rehydration processes of cement paste
exposed to high temperature environments’, Journal of materials science 39(9), 3015–3024.
Analla, M. (1998), ‘Model validation through the linear regression fit to actual versus predicted
values’, Agricultural Systems 57(1), 115–119.
Anstis, G., Chantikul, P., Lawn, B. R. and Marshall, D. (1981), ‘A critical evaluation of inden-
tation techniques for measuring fracture toughness: I, direct crack measurements’, Journal
of the American Ceramic Society 64(9), 533–538.
ANSYS R○ (2012), ‘Ansys mechanical apdl theory reference’, Documentation for Ansys .
ANSYS R○ (2015), ‘Release 16.2’, Documentation for Ansys .
ASTM Standard C109 (2011), Standard test method for compressive strength of hydraulic
cement mortars (using 2-in. or [50-mm] cube specimens), Standard, ASTM International
West Conshohocken, PA.
ASTM Standard C140 (2015), Standard test methods for sampling and testing concrete masonry
units and related units, Standard, ASTM International West Conshohocken, PA.
ASTM Standard C150 (2015), Standard specification of portland cement, Standard, ASTM
International West Conshohocken, PA.
ASTM Standard C1609 (2012), Standar test method for flexural performance of fibre-reinforced
concrete (using beam with third-point loading), Standard, ASTM International West Con-
shohocken, PA.
ASTM Standard C618 (2012), Specification for coal fly ash and raw or calcined natural pozzolan
for use in concrete, Standard, ASTM International West Conshohocken, PA.
ASTM Standard C642 (2013), Standard test method for density, absorption, and voids in
hardened concrete, Standard, ASTM International West Conshohocken, PA.
Australia Standard (2009), Supplementary cementitious materials for use with portland and
blended cement, Standard, Standard Australia.
Australia Standard (2010), General purpose and blended cements, Standard, Standard Aus-
tralia.
Aydın, S. (2008), ‘Development of a high-temperature-resistant mortar by using slag and
pumice’, Fire safety journal 43(8), 610–617.
Berry, E., Hemmings, R. and Cornelius, B. (1990), ‘Mechanisms of hydration reactions in high
volume fly ash pastes and mortars’, Cement and Concrete Composites 12(4), 253–261.
204
Bhattacharya, A. and Nix, W. (1991), ‘Finite element analysis of cone indentation’, Interna-
tional Journal of Solids and Structures 27(8), 1047–1058.
Bobko, C. and Ulm, F.-J. (2008), ‘The nano-mechanical morphology of shale’, Mechanics of
Materials 40(4), 318–337.
Budiansky, B. and Cui, Y. L. (1994), ‘On the tensile strength of a fiberreinforced ceramic com-
posite containing a crack-like flaw’, Journal of the Mechanics and Physics of Solids 42(1), 1–
19.
Bye, G. C. (1999), Portland cement: composition, production and properties, Thomas Telford.
Caijun, S. and Della, R. (2006), ‘Alkali-activated cements and concretes’, London and New
York: Taylor & Francis .
Cao, Y.-P., Ji, X.-Y. and Feng, X.-Q. (2010), ‘Geometry independence of the normalized relax-
ation functions of viscoelastic materials in indentation’, Philosophical Magazine 90(12), 1639–
1655.
Cao, Z. and Zhang, X. (2008), ‘Nanoindentation stress–strain curves of plasma-enhanced chem-
ical vapor deposited silicon oxide thin films’, Thin Solid Films 516(8), 1941–1951.
Chan, S. Y., Peng, G.-f. and Chan, J. K. (1996), ‘Comparison between high strength con-
crete and normal strength concrete subjected to high temperature’, Materials and Structures
29(10), 616–619.
Chen, J. and Bull, S. (2007), ‘Indentation fracture and toughness assessment for thin optical
coatings on glass’, Journal of Physics D: Applied Physics 40(18), 5401.
Chen, X., Beyerlein, I. J. and Brinson, L. C. (2009a), ‘Curved-fiber pull-out model for nanocom-
posites. part 1: Bonded stage formulation’, Mechanics of Materials 41(3), 279–292.
Chen, X., Beyerlein, I. J. and Brinson, L. C. (2009b), ‘Curved-fiber pull-out model for nanocom-
posites. part 2: Interfacial debonding and sliding’, Mechanics of Materials 41(3), 293–307.
Chen, X., Meawad, A. and Struble, L. J. (2014), ‘Method to stop geopolymer reaction’, Journal
of the American Ceramic Society 97(10), 3270–3275.
Cheng, Y.-T. and Cheng, C.-M. (2004), ‘Scaling, dimensional analysis, and indentation mea-
surements’, Materials Science and Engineering: R: Reports 44(4), 91–149.
Chindaprasirt, P., Jaturapitakkul, C. and Sinsiri, T. (2005), ‘Effect of fly ash fineness on com-
pressive strength and pore size of blended cement paste’, Cement and Concrete Composites
27(4), 425–428.
Constantinides, G., Chandran, K. R., Ulm, F.-J. and Van Vliet, K. (2006), ‘Grid indentation
analysis of composite microstructure and mechanics: principles and validation’, Materials
Science and Engineering: A 430(1), 189–202.
Constantinides, G. and Ulm, F.-J. (2004), ‘The effect of two types of csh on the elasticity
of cement-based materials: Results from nanoindentation and micromechanical modeling’,
Cement and concrete research 34(1), 67–80.
Constantinides, G. and Ulm, F.-J. (2007), ‘The nanogranular nature of csh’, Journal of the
Mechanics and Physics of Solids 55(1), 64–90.
205
Criado, M., Fernandez-Jimenez, A., Palomo, A., Sobrados, I. and Sanz, J. (2008), ‘Effect of
the sio 2/na 2 o ratio on the alkali activation of fly ash. part ii: 29 si mas-nmr survey’,
Microporous and Mesoporous Materials 109(1), 525–534.
Criado, M., Palomo, A. and Fernandez-Jimenez, A. (2005), ‘Alkali activation of fly ashes. part
1: Effect of curing conditions on the carbonation of the reaction products’, Fuel 84(16), 2048–
2054.
Davidovits, J. (1994a), ‘High-alkali cements for 21st century concretes’, ACI Special Publication
144.
Davidovits, J. (1994b), Properties of geopolymer cements, in ‘First international conference on
alkaline cements and concretes’, Vol. 1, pp. 131–149.
de Normalisation, C. E. (2005), ‘Eurocode 4: Design of composite steel and concrete structures.
part 1-2: General rules-structural fire design’, CEN ENV .
de Normalisation, C. E. (2015), ‘Eurocode 2: Design of concrete structures - part 1-1: General
rules and rules for building’, CEN ENV .
Demirel, B. and Kelestemur, O. (2010), ‘Effect of elevated temperature on the mechanical
properties of concrete produced with finely ground pumice and silica fume’, Fire Safety
Journal 45(6), 385–391.
Diamond, S. (2000), ‘Mercury porosimetry: an inappropriate method for the measure-
ment of pore size distributions in cement-based materials’, Cement and concrete research
30(10), 1517–1525.
Dormieux, L., Kondo, D. and Ulm, F.-J. (2006), Microporomechanics, John Wiley & Sons.
Dukino, R. D. and Swain, M. V. (1992), ‘Comparative measurement of indentation fracture
toughness with berkovich and vickers indenters’, Journal of the American Ceramic Society
75(12), 3299–3304.
Duxson, P., Fernandez-Jimenez, A., Provis, J. L., Lukey, G. C., Palomo, A. and Van Deventer,
J. (2007), ‘Geopolymer technology: the current state of the art’, Journal of Materials Science
42(9), 2917–2933.
Duxson, P., Provis, J. L., Lukey, G. C. and Van Deventer, J. S. (2007), ‘The role of inorganic
polymer technology in the development of green concrete’, Cement and Concrete Research
37(12), 1590–1597.
Fernandez-Jimenez, A., De La Torre, A., Palomo, A., Lopez-Olmo, G., Alonso, M. and Aranda,
M. (2006), ‘Quantitative determination of phases in the alkaline activation of fly ash. part ii:
Degree of reaction’, Fuel 85(14), 1960–1969.
Fernandez-Jimenez, A. and Palomo, A. (2003), ‘Characterisation of fly ashes. potential reactiv-
ity as alkaline cements’, Fuel 82(18), 2259–2265.
Fernandez-Jimenez, A. and Palomo, A. (2005), ‘Composition and microstructure of alkali acti-
vated fly ash binder: effect of the activator’, Cement and Concrete Research 35(10), 1984–
1992.
Ferry, J. D. (1980), Viscoelastic properties of polymers, John Wiley & Sons.
206
Field, J., Swain, M. and Dukino, R. (2003), ‘Determination of fracture toughness from the
extra penetration produced by indentation-induced pop-in’, Journal of materials research
18(06), 1412–1419.
Findley, W. N. and Davis, F. A. (2013), Creep and relaxation of nonlinear viscoelastic materials,
Courier Corporation.
Fischer-Cripps, A. (1999), ‘The hertzian contact surface’, Journal of materials science
34(1), 129–137.
Fischer-Cripps, A. (2004), ‘A simple phenomenological approach to nanoindentation creep’,
Materials Science and Engineering: A 385(1), 74–82.
Fischer-Cripps, A. C. (2011), Nanoindentation, Springer.
Fischer-Cripps, A. C. and Mustafaev, I. (2000), Introduction to contact mechanics, Springer.
Flatt, R. J., Roussel, N. and Cheeseman, C. R. (2012), ‘Concrete: An eco material that needs
to be improved’, Journal of the European Ceramic Society 32(11), 2787–2798.
French, D. and Smitham, J. (2007), Fly ash characteristics and feed coal properties, QCAT
Technology Transfer Centre.
Gaillard, Y., Tromas, C. and Woirgard, J. (2003), ‘Pop-in phenomenon in mgo and lif: obser-
vation of dislocation structures’, Philosophical magazine letters 83(9), 553–561.
Garcia-Lodeiro, I., Palomo, A., Fernandez-Jimenez, A. and Macphee, D. (2011), ‘Compatibility
studies between nash and cash gels. study in the ternary diagram na 2 o–cao–al 2 o 3–sio 2–h
2 o’, Cement and Concrete Research 41(9), 923–931.
Glukhovsky, V., Rostovskaja, G. and Rumyna, G. (1980), High strength slag-alkaline cements,
in ‘Proceedings of the 7th international congress on the chemistry of cement, Paris’.
Guide to the use of fly ash in concrete in Australia (2009), Report, Ash Development Association
of Australia.
Handoo, S., Agarwal, S. and Agarwal, S. (2002), ‘Physicochemical, mineralogical, and morpho-
logical characteristics of concrete exposed to elevated temperatures’, Cement and Concrete
Research 32(7), 1009–1018.
Hardijito, D. and Rangan, B. (2005), Development and properties of low-calcuim fly ash-based
geopolymer concrete, Report, Faculty of Engineering.
Hardjito, D., Wallah, S. E., Sumajouw, D. M. and Rangan, B. V. (2004), ‘On the development
of fly ash-based geopolymer concrete’, ACI materials journal 101(6).
Heidrich, C. and Woodhead, A. (2005), Prepared for the benchmarking module: Sustainablility
capcity building program, Report, Pavement Recycling and Stabilisation Association.
Helmuth, R. (1983), ‘Some questions concerning astm standards and methods of testing fly ash
for use with portland cement’, Cement, concrete and aggregates 5(2), 103–110.
Herbert, E., Pharr, G., Oliver, W., Lucas, B. and Hay, J. (2001), ‘On the measurement of
stressstrain curves by spherical indentation’, Thin solid films 398, 331–335.
207
Herrera-Franco, P. and Drzal, L. (1992), ‘Comparison of methods for the measurement of
fibre/matrix adhesion in composites’, Composites 23(1), 2–27.
Hsueh, C.-H. (1990), ‘Interfacial debonding and fiber pull-out stresses of fiber-reinforced com-
posites’, Materials Science and Engineering: A 123(1), 1–11.
ISO 14577-1 (2002), Metallic materials-instrumented indentation test for hardness and materials
parameters-part 1: test method., Standard, International Organization for Standardization
Geneva, Switzerland.
Jennings, H. M. (2000), ‘A model for the microstructure of calcium silicate hydrate in cement
paste’, Cement and Concrete Research 30(1), 101–116.
Jennings, H. M., Thomas, J. J., Gevrenov, J. S., Constantinides, G. and Ulm, F.-J. (2007),
‘A multi-technique investigation of the nanoporosity of cement paste’, Cement and Concrete
Research 37(3), 329–336.
Joseph, D. (2011), ‘Geopolymer chemistry & applications’.
Kanda, T., Lin, Z. and Li, V. C. (2000), ‘Tensile stress-strain modeling of pseudostrain hard-
ening cementitious composites’, Journal of Materials in Civil Engineering 12(2), 147–156.
Kanit, T., Forest, S., Galliet, I., Mounoury, V. and Jeulin, D. (2003), ‘Determination of the
size of the representative volume element for random composites: statistical and numerical
approach’, International Journal of solids and structures 40(13), 3647–3679.
Kim, D. J., Park, S. H., Ryu, G. S. and Koh, K. T. (2011), ‘Comparative flexural behav-
ior of hybrid ultra high performance fiber reinforced concrete with different macro fibers’,
Construction and Building Materials 25(11), 4144–4155.
Kim, J.-K. and Yi, S.-T. (2002), ‘Application of size effect to compressive strength of concrete
members’, Sadhana 27(4), 467–484.
Kong, D. L. and Sanjayan, J. G. (2008), ‘Damage behavior of geopolymer composites exposed
to elevated temperatures’, Cement and Concrete Composites 30(10), 986–991.
Kong, D. L. and Sanjayan, J. G. (2010), ‘Effect of elevated temperatures on geopolymer paste,
mortar and concrete’, Cement and Concrete Research 40(2), 334–339.
Kullaa, J. (1996), ‘Dimensional analysis of bond modulus in fiber pullout’, Journal of Structural
Engineering 122(7), 783–787.
Kumar, S., Kumar, R., Alex, T., Bandopadhyay, A. and Mehrotra, S. (2005), Effect of mechan-
ically activated fly ash on the properties of geopolymer cement, in ‘Proceedings of the 4th
World Congress on Geopolymer’, pp. 113–116.
Labuz, J. F. and Zang, A. (2015), Mohr–coulomb failure criterion, in ‘The ISRM Suggested
Methods for Rock Characterization, Testing and Monitoring: 2007-2014’, Springer, pp. 227–
231.
Laugier, M. (1987), ‘Palmqvist indentation toughness in wc-co composites’, Journal of materials
science letters 6(8), 897–900.
208
Lawn, B. R., Evans, A. and Marshall, D. (1980), ‘Elastic/plastic indentation damage in ceram-
ics: the median/radial crack system’, Journal of the American Ceramic Society 63(910), 574–
581.
Li, V. C. (1997), ‘Engineered cementitious composites (ecc)–tailored composites through micro-
mechanical modelling’, Fiber Reinforced Concrete: Present and the Future, Canadian Society
for Civil Engineering pp. 64–97.
Li, V. C. and Leung, C. K. (1992), ‘Steady-state and multiple cracking of short random fiber
composites’, Journal of Engineering Mechanics 118(11), 2246–2264.
Li, V. C., Mishra, D. K. and Wu, H.-C. (1995), ‘Matrix design for pseudo-strain-hardening fibre
reinforced cementitious composites’, Materials and Structures 28(10), 586–595.
Li, V. C., Wang, Y. and Backer, S. (1990), ‘Effect of inclining angle, bundling and surface
treatment on synthetic fibre pull-out from a cement matrix’, Composites 21(2), 132–140.
Lichinchi, M., Lenardi, C., Haupt, J. and Vitali, R. (1998), ‘Simulation of berkovich nanoinden-
tation experiments on thin films using finite element method’, Thin solid films 312(1), 240–
248.
Ling, F. F. (2011), Nanoindentation, third edition edn, Springer, New York.
Litvan, G. G. (1986), Mechanism of cement paste degradation due to chemical and physical
processes, National Research Council Canada, Institute for Research in Construction.
Lloyd, R. R., Provis, J. L. and Van Deventer, J. S. (2010), ‘Pore solution composition and alkali
diffusion in inorganic polymer cement’, Cement and Concrete Research 40(9), 1386–1392.
Marques, S. P. and Creus, G. J. (2012), Rheological Models: Integral and Differential Represen-
tations, Springer, pp. 11–21.
Martinez, E., Romero, J., Lousa, A. and Esteve, J. (2003), ‘Nanoindentation stressstrain curves
as a method for thin-film complete mechanical characterization: application to nanometric
crn/cr multilayer coatings’, Applied Physics A 77(3-4), 419–427.
MATLAB (2014), R2014a, The MathWorks Inc., Natick, Massachusetts.
Mendes, A., Sanjayan, J. and Collins, F. (2008), ‘Phase transformations and mechanical
strength of opc/slag pastes submitted to high temperatures’, Materials and structures
41(2), 345–350.
Miller, M., Bobko, C., Vandamme, M. and Ulm, F.-J. (2008), ‘Surface roughness criteria for
cement paste nanoindentation’, Cement and Concrete Research 38(4), 467–476.
Morsy, M., Rashad, A. and Shebl, S. (2008), ‘Effect of elevated temperature on compressive
strength of blended cement mortar’, Build Res J 56(2-3), 173–185.
Morsy, M., Shebl, S. and Rashad, A. (2009), ‘Effect of fire on microstructure and mechan-
ical properties of blended cement pastes containing metakaolin and silica fume’, Silicates
Industriels 74(3), 59.
Naaman, A. E. (1972), A statistical theory of strength for fiber reinforced concrete, PhD thesis,
Massachusetts Institute of Technology.
209
Naaman, A. E., Moavenzadeh, F. and McGarry, F. J. (1974), ‘Probabilistic analysis of fiber-
reinforced concrete’, Journal of the Engineering Mechanics Division 100(2), 397–413.
Naaman, A. E., Namur, G. G., Alwan, J. M. and Najm, H. S. (1991), ‘Fiber pullout and bond
slip. i: Analytical study’, Journal of Structural Engineering 117(9), 2769–2790.
Naaman, A. E. and Reinhardt, H. (2008), ‘High performance fiber reinforced cement com-
posites’, Naaman AE. High-performance construction materials: science and applications.
Singapore: World Scientific Publishing pp. 91–153.
Nashif, A. D., Jones, D. I. and Henderson, J. P. (1985), Vibration damping, John Wiley & Sons.
Nemecek, J., Smilauer, V. and Kopecky, L. (2011), ‘Nanoindentation characteristics of alkali-
activated aluminosilicate materials’, Cement and Concrete Composites 33(2), 163–170.
Neville, A. M. (2011), Properties of concrete.
Oliver, W. C. and Pharr, G. M. (1992), ‘An improved technique for determining hardness and
elastic modulus using load and displacement sensing indentation experiments’, Journal of
materials research 7(06), 1564–1583.
Oliver, W. C. and Pharr, G. M. (2004), ‘Measurement of hardness and elastic modulus by instru-
mented indentation: Advances in understanding and refinements to methodology’, Journal
of materials research 19(01), 3–20.
Ozbay, E., Oztas, A., Baykasoglu, A. and Ozbebek, H. (2009), ‘Investigating mix proportions of
high strength self compacting concrete by using taguchi method’, Construction and Building
Materials 23(2), 694–702.
Pacheco-Torgal, F., Castro-Gomes, J. and Jalali, S. (2008), ‘Alkali-activated binders: A review:
Part 1. historical background, terminology, reaction mechanisms and hydration products’,
Construction and Building Materials 22(7), 1305–1314.
Palomo, A., Grutzeck, M. and Blanco, M. (1999), ‘Alkali-activated fly ashes: a cement for the
future’, Cement and concrete research 29(8), 1323–1329.
Papadakis, V. G. (1999), ‘Effect of fly ash on portland cement systems: Part i. low-calcium fly
ash’, Cement and Concrete Research 29(11), 1727–1736.
Papadakis, V. G. (2000), ‘Effect of fly ash on portland cement systems: Part ii. high-calcium
fly ash’, Cement and Concrete Research 30(10), 1647–1654.
Park, G.-J. (2007), ‘Design of experiments’, Analytic Methods for Design Practice pp. 309–391.
Peiwei, G., Min, D. and Naiqian, F. (2001), ‘The influence of superplasticizer and superfine
mineral powder on the flexibility, strength and durability of hpc’, Cement and Concrete
Research 31(5), 703–706.
Pichler, C. and Lackner, R. (2008), ‘A multiscale creep model as basis for simulation of earlyage
concrete behavior’, Computers and Concrete 5(4), 295–328.
Pipilikaki, P. and Beazi-Katsioti, M. (2009), ‘The assessment of porosity and pore size distribu-
tion of limestone portland cement pastes’, Construction and Building Materials 23(5), 1966–
1970.
210
Pipkin, A. C. (2012), Lectures on viscoelasticity theory, Vol. 7, Springer Science & Business
Media.
Poon, C., Lam, L. and Wong, Y. (2000), ‘A study on high strength concrete prepared with
large volumes of low calcium fly ash’, Cement and Concrete Research 30(3), 447–455.
Poon, C.-S., Azhar, S., Anson, M. and Wong, Y.-L. (2001), ‘Comparison of the strength and
durability performance of normal-and high-strength pozzolanic concretes at elevated temper-
atures’, Cement and Concrete Research 31(9), 1291–1300.
Powers, T. C. (1958), ‘Structure and physical properties of hardened portland cement paste’,
Journal of the American Ceramic Society 41(1), 1–6.
Provis, J. L. and Van Deventer, J. S. J. (2009), Geopolymers: structures, processing, properties
and industrial applications, Elsevier.
Ramezanianpour, A. A. (2014), Cement replacement materials, Springer.
Rangan, B. (2008), Fly ash-based geopolymer concrete, Report, Faculty of Engineering.
Ranjit, K. (1990), ‘A primer on the taguchi method, competitive manufacturing series’.
Ross, P. J. (1996), ‘Taguchi techniques for quality engineering: loss function, orthogonal exper-
iments, parameter and tolerance design’.
Roy, D. M. (1999), ‘Alkali-activated cements opportunities and challenges’, Cement and Con-
crete Research 29(2), 249–254.
Roy, R. K. (2010), A primer on the Taguchi method, Society of Manufacturing Engineers.
Sakharova, N., Fernandes, J., Antunes, J. and Oliveira, M. (2009), ‘Comparison between
berkovich, vickers and conical indentation tests: A three-dimensional numerical simulation
study’, International Journal of Solids and Structures 46(5), 1095–1104.
Scrivener, K. L. (2004), ‘Backscattered electron imaging of cementitious microstructures: un-
derstanding and quantification’, Cement and Concrete Composites 26(8), 935–945.
Scrivener, K. L. and Nonat, A. (2011), ‘Hydration of cementitious materials, present and future’,
Cement and concrete research 41(7), 651–665.
Seber, G. A. and Lee, A. J. (2012), Linear regression analysis, Vol. 936, John Wiley & Sons.
Shah, S. P. (1989), Fracture toughness of cement-based materials, in ‘Fracture of Concrete and
Rock’, Springer, pp. 1–12.
Shi, C. (1992), Activation of natural pozzolans, fly ashes and blast furnace slag, Civil Engineer-
ing, University of Calgary.
Shi, C. (1996), ‘Strength, pore structure and permeability of alkali-activated slag mortars’,
Cement and Concrete Research 26(12), 1789–1799.
Shi, C. and Day, R. L. (1993a), ‘Acceleration of strength gain of lime-pozzolan cements by
thermal activation’, Cement and Concrete Research 23(4), 824–832.
Shi, C. and Day, R. L. (1993b), ‘Chemical activation of blended cements made with lime and
natural pozzolans’, Cement and Concrete Research 23(6), 1389–1396.
211
Shi, C. and Day, R. L. (2000), ‘Pozzolanic reaction in the presence of chemical activators: Part
iireaction products and mechanism’, Cement and Concrete Research 30(4), 607–613.
Shi, C. and Fernandez-Jimenez, A. (2006), ‘Stabilization/solidification of hazardous and ra-
dioactive wastes with alkali-activated cements’, Journal of hazardous materials 137(3), 1656–
1663.
Shi, C., Jimenez, A. F. and Palomo, A. (2011), ‘New cements for the 21st century: the pursuit
of an alternative to portland cement’, Cement and Concrete Research 41(7), 750–763.
Siddique, R. (2004), ‘Performance characteristics of high-volume class f fly ash concrete’, Ce-
ment and Concrete Research 34(3), 487–493.
Sneddon, I. N. (1965), ‘The relation between load and penetration in the axisymmetric boussi-
nesq problem for a punch of arbitrary profile’, International Journal of Engineering Science
3(1), 47–57.
Soranakom, C. and Mobasher, B. (2008), ‘Correlation of tensile and flexural responses of
strain softening and strain hardening cement composites’, Cement and concrete Composites
30(6), 465–477.
Srinivasan, C., Narasimhan, N. L. and Ilango, S. (2003), ‘Development of rapid-set high-strength
cement using statistical experimental design’, Cement and Concrete Research 33(9), 1287–
1292.
Stauss, S., Schwaller, P., Bucaille, J.-L., Rabe, R., Rohr, L., Michler, J. and Blank, E. (2003),
‘Determining the stress–strain behaviour of small devices by nanoindentation in combination
with inverse methods’, Microelectronic Engineering 67, 818–825.
Stein, E. M. (1956), ‘Interpolation of linear operators’, Transactions of the American Mathe-
matical Society pp. 482–492.
Strategies, E. (2007), ‘Review of co2-e emissions from concrete versus timber sleepers’.
Stutzman, P. E. (1996), Guide for X-ray powder diffraction analysis of Portland cement and
clinker, US Department of Commerce, Technology Administration, National Institute of Stan-
dards and Technology, Office of Applied Economics, Building and Fire Research Laboratory.
Stutzman, P. E. (2001), ‘Scanning electron microscopy in concrete petrography’, National In-
stitute of Standards and Technology 2.
Taha, M. R., Soliman, E., Sheyka, M., Reinhardt, A. and Al-Haik, M. (2010), ‘Fracture tough-
ness of hydrated cement paste using nanoindentation’.
Tanyildizi, H. (2013), ‘Variance analysis of crack characteristics of structural lightweight con-
crete containing silica fume exposed to high temperature’, Construction and Building Mate-
rials 47, 1154–1159.
Taylor, H. F. (1997), Cement chemistry, Thomas Telford.
Tennis, P. D. and Jennings, H. M. (2000), ‘A model for two types of calcium silicate hydrate in
the microstructure of portland cement pastes’, Cement and Concrete Research 30(6), 855–
863.
212
Thomas, J. J., Jennings, H. M. and Allen, A. J. (1998), ‘The surface area of cement paste as
measured by neutron scattering: evidence for two csh morphologies’, Cement and Concrete
Research 28(6), 897–905.
Turner, L. K. and Collins, F. G. (2013), ‘Carbon dioxide equivalent (co 2-e) emissions: a com-
parison between geopolymer and opc cement concrete’, Construction and Building Materials
43, 125–130.
Ulm, F.-J., Vandamme, M., Bobko, C., Alberto Ortega, J., Tai, K. and Ortiz, C. (2007),
‘Statistical indentation techniques for hydrated nanocomposites: concrete, bone, and shale’,
Journal of the American Ceramic Society 90(9), 2677–2692.
U.S. Environmental Protection agency (2007), ‘Waste-resource conservation-industraul material
recycling’.
Van den Heede, P. and De Belie, N. (2012), ‘Environmental impact and life cycle assessment
(lca) of traditional and greenconcretes: literature review and theoretical calculations’, Cement
and Concrete Composites 34(4), 431–442.
Vandamme, M., Tweedie, C. A., Constantinides, G., Ulm, F.-J. and Van Vliet, K. J. (2012),
‘Quantifying plasticity-independent creep compliance and relaxation of viscoelastoplastic ma-
terials under contact loading’, Journal of Materials Research 27(01), 302–312.
Vandamme, M. and Ulm, F.-J. (2006), ‘Viscoelastic solutions for conical indentation’, Interna-
tional Journal of solids and structures 43(10), 3142–3165.
Vandamme, M. and Ulm, F.-J. (2009), ‘Nanogranular origin of concrete creep’, Proceedings of
the National Academy of Sciences 106(26), 10552–10557.
Vandamme, M. and Ulm, F.-J. (2013), ‘Nanoindentation investigation of creep properties of
calcium silicate hydrates’, Cement and Concrete Research 52, 38–52.
Velez, K., Maximilien, S., Damidot, D., Fantozzi, G. and Sorrentino, F. (2001), ‘Determination
by nanoindentation of elastic modulus and hardness of pure constituents of portland cement
clinker’, Cement and Concrete Research 31(4), 555–561.
Wallah, S. and Rangan, B. V. (2006), ‘Low-calcium fly ash-based geopolymer concrete: long-
term properties’, Res. Report-GC2, Curtin University, Australia. pp pp. 76–80.
Ward, R. and Li, V. C. (1991), ‘Dependence of flexural behaviour of fibre reinforced mortar on
material fracture resistance and beam size’, Construction and Building Materials 5(3), 151–
161.
Warner, R. F., Rangan, B., Hall, A. and Faulkes, K. (1998), Concrete structures.
What is Lime? (2005), Report, Pavement Recycling and Stabilisation Association.
Xu, H. and Van Deventer, J. (2000), ‘The geopolymerisation of alumino-silicate minerals’,
International Journal of Mineral Processing 59(3), 247–266.
Xu, Y., Jiang, L., Xu, J. and Li, Y. (2012), ‘Mechanical properties of expanded polystyrene
lightweight aggregate concrete and brick’, Construction and Building Materials 27(1), 32–38.
Xu, Y., Wong, Y., Poon, C. and Anson, M. (2001), ‘Impact of high temperature on pfa concrete’,
Cement and Concrete Research 31(7), 1065–1073.
213
Zaoui, A. (2002), ‘Continuum micromechanics: survey’, Journal of Engineering Mechanics
128(8), 808–816.
Zhan, Y. and Meschke, G. (2014), ‘Analytical model for the pullout behavior of straight and
hooked-end steel fibers’, Journal of Engineering Mechanics 140(12), 04014091.
Zhang, J. and Scherer, G. W. (2011), ‘Comparison of methods for arresting hydration of cement’,
Cement and Concrete Research 41(10), 1024–1036.
214