fulltext (22)

Title Properties of mortars with binary and ternary blendedcementitious materials

Advisor(s) Kwan, AKH

Author(s) Li, Yan; 李彦

Citation

Issued Date 2012

URL http://hdl.handle.net/10722/173879

Rights The author retains all proprietary rights, (such as patent rights)and the right to use in future works.

PROPERTIES OF MORTARS WITH BINARY AND

TERNARY BLENDED CEMENTITIOUS MATERIALS

LI YAN

M.PHIL. THESIS

THE UNIVERSITY OF HONG KONG

2012



by

LI YAN

(李彦李彦李彦李彦)

B.Eng. (Civil Engineering),

Huazhong University of Science and Technology

A thesis submitted in partial fulfilment of the requirements for

the Degree of Master of Philosophy

at The University of Hong Kong

September 2012

Abstract of thesis entitled



Submitted by

LI Yan

for the Degree of Master of Philosophy

at The University of Hong Kong

in September 2012

During the past few decades, concrete technology has been developing

rapidly followed with huge popularity of high-performance concrete (HPC).

However, the mix design for HPC still remains a major challenge due to the wide

adoption of mineral and chemical admixtures, the effects of which are rather

complicated and not yet fully understood. To resolve this issue, this thesis

presents a comprehensive experimental study focused on the physical effects of

some supplementary cementitious materials (SCM) on the fresh and hardened

properties of mortar. Based on the experimental results, some fundamental

parameters governing the performance of mortar were investigated.

It has been postulated by some researches that increasing the packing

density of the particle system would improve the rheology and strength of

concrete. Through adding SCM finer than cement to increase the packing density,

the voids between solid particles will be reduced so that more excess water can be

released to provide better lubrication. Through adding two kinds of SCMs with

different fineness, the packing density will be further enhanced by the successive

filling action. In this study, a wet packing method, which is newly developed at

the University of Hong Kong, was used to directly measure the packing densities

of mortars with binary and ternary blended cementitious materials. The filling

effect and successive filling action were both quantified through the packing

density results.

The study revealed that the addition of fine SCM will, not only increase

the packing density, but also increase the solid surface area, which will have

negative effect on the rheology of mortar. To combine the effects of water content,

packing density and solid surface area together, we proposed a new parameter

called water film thickness (WFT), defined as the average thickness of water films

coating the solid particles and evaluated as the excess water to solid surface area

ratio. The results demonstrated that the WFT plays a key role in controlling the

rheology and strength of mortar. Hence, it is the WFT, rather than the packing

density, that should be maximized at given water content in the mix design of

HPC. The addition of fine SCM will increase both the excess water content and

solid surface area. The effects on the both sides can be quantified by the WFT no

matter how complex the cementitious system is. Therefore, the WFT could be

used as an effective indicator to adjust the SCM content. Joint addition of fine

SCM at different level finer than cement to make a ternary cementitious system

can effectively increase the packing density without excessively increasing the

solid surface area. As a result, the ternary cementitious system has higher

effectiveness than the binary cementitious system in improving the performance

of mortar.

(446words)

Dedicated to My Beloved Family

- i -

DECLARATION

I declare that this thesis represents my own work, except where due

acknowledgement is made, and that it has not been previously included in a thesis,

dissertation or report submitted to this University or to any other institution for a

degree, diploma or other qualification.

Signed

LI Yan

Date

- ii -

ACKNOWLEDGEMENTS

The author is particularly grateful to his supervisor, Professor A.K.H.

Kwan, for his invaluable advice and guidance throughout the study. The author

learnt and grew a lot during the many times of talks with him. His insightful

inspiration, far vision and generous encouragement benefited the author greatly

during the past two years and will remain precious wealth in the author’s life.

Without his unconditional support, the author’s postgraduate study will never be

achievable.

The author would like to thank all the staff members in the Department of

Civil Engineering, the undergraduate student research assistants and final year

project students for their assistance in the experimental works. Sincere thanks are

also to the authors’ fellow postgraduate colleagues in the University of Hong

Kong, including J.J. Chen, H.W. Zhang, C.C.Y. Leung, L.G. Li, M. McKinley,

K.W. Chan and V. Wong, for their kind collaboration. The author would like to

thank the Research Grants Council of the Hong Kong Special Administrative

Region for the studentship it offered.

The last but not the least, words cannot express the author’s immense

gratitude to his beloved parents and girlfriend for their understanding and support

throughout the author’s postgraduate study.

- iii -

CONTENTS

DECLARATION i

ACKNOWLEDGEMENTS ii

TABLE OF CONTENTS iii

LIST OF TABLES vi

LIST OF FIGURES vii

LIST OF SYMBOLS viii

LIST OF ABBREVIATIONS ix

CHAPTER 1: INTRODUCTION

1.1 Background 1

1.2 Development of High-Performance Concrete 3

1.3 Three-Tier Mix Design Approach 6

1.4 Development of Green Concrete 7

1.5 Objectives of Research 7

1.6 Organization of Thesis 8

CHAPTER 2: EXPERIMENTAL PROGRAM

2.1 Measurement of Packing Density 10

2.2 Determination of Water Film Thickness 11

2.3 Measurement of Flowability 12

2.4 Measurement of Rheological properties 13

2.5 Measurement of Adhesiveness 15

2.6 Measurement of Strength 15

CHAPTER 3: BINARY CEMENTITIOUS SYSTEM CONTAINING FLY

ASH MICROSPHERE

3.1 Introduction 19

3.2 Materials 22

3.3 Experimental Program 23

3.4 Test Methods 24

- iv -

3.5 Experimental Results 25

3.5.1 Packing density and voids ratio 25

3.5.2 Water film thickness 26

3.5.3 Flow spread and flow rate 27

3.5.4 Yield stress and apparent viscosity 28

3.5.5 Correlation between flowability and rheological properties 28

3.5.6 Adhesiveness 29

3.5.7 Cube strength 30

3.6 Concurrent Strength-Flowability Performance 32

3.7 Conclusions 32

CHAPTER 4: TERNARY CEMENTITIOUS SYSTEM CONTAINING FLY

ASH MICROSPHERE AND CONDENSED SILICA FUME

4.1 Introduction 47

4.2 Materials 50


4.4 Test Methods 52


4.5.1 Packing density and voids ratio 53

4.5.2 Water film thickness 54

4.5.3 Flow spread 56

4.5.4 Flow rate 57

4.5.5 Yield stress 58

4.5.6 Apparent viscosity 59



4.6 Concurrent Strength-Flowability Performance 62

4.7 Conclusions 62

CHAPTER 5: TERNARY CEMENTITIOUS SYSTEM CONTAINING

SUPERFINE CEMENT AND CONDENSED SILICA FUME

5.1 Introduction 77

5.2 Materials 80


- v -

5.4 Test Methods 82


5.5.1 Packing density and water film thickness 83

5.5.2 Flow spread and flow rate 85

5.5.3 Yield stress and apparent viscosity 86



5.6 Roles of Water Film Thickness 87

5.6.1 Effects of WFT on flow spread and flow rate 88

5.6.2 Effects of WFT on yield stress and apparent viscosity 89

5.6.3 Effect of WFT on adhesiveness 90

5.6.4 Effect of WFT on cube strength 91

5.7 Conclusions 91

CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS

6.1 Summary of Study 104

6.1.1 Role of packing density 105

6.1.2 Role of water film thickness 105

6.1.3 Effects of fly ash microsphere 106

6.1.4 Effects of superfine cement 107

6.1.5 Effects of condensed silica fume 107

6.1.6 Effects of ternary blended system 108

6.2 Conclusions 108

6.3 Recommendations for Future Work 111

REFERENCES 112

LIST OF PUBLICATIONS DURING CANDIDATURE 118

- vi -

List of Tables

Table Page

3.1 Chemical compositions of OPC and FAM 34

3.2 Mix proportions, packing density and WFT 35

3.3 Flowability, rheological properties, adhesiveness and strength 36





- vii -

List of Figures

Figure Page

2.1 (a) Mini slump cone; (b) Mini V-funnel 16

2.2 Rheometer, shear vane and container 17

2.3 Details of apparatus for stone rod adhesion test 18

3.1 Particle size distributions of FAM, OPC and fine aggregate 37

3.2 Scanning electron microscope image of FAM 38

3.3 Variations of packing density and voids ratio with FAM content 39

3.4 Variation of WFT with W/S ratio at different FAM contents 40

3.5 Flowability versus WFT 41

3.6 Rheological properties versus WFT 42

3.7 Correlation between flowability and rheological properties 43

3.8 Adhesiveness versus WFT 44

3.9 Cube strength versus W/CM ratio and WFT 45

3.10 Concurrent strength and flowability performance 46

4.1 Particle size distributions of CSF, FAM, OPC and fine aggregate 68

4.2 Variation of WFT with W/S ratio at different FAM and CSF contents 69

4.3 Flow spread versus W/S ratio and WFT 70

4.4 Flow rate versus W/S ratio and WFT 71

4.5 Yield stress versus W/S ratio and WFT 72

4.6 Apparent viscosity versus W/S ratio and WFT 73

4.7 Adhesiveness versus W/S ratio and WFT 74

4.8 Cube strength versus W/S ratio and WFT 75

4.9 Concurrent cube strength and flowability performance 76

5.1 Particle size distributions of CSF, SFC, OPC and fine aggregate 95

5.2 Variation of WFT with W/S ratio at different SFC and CSF contents 96

5.3 Flowability versus W/S ratio 97

5.4 Rheological properties versus W/S ratio 98

5.5 Adhesiveness versus W/S ratio 99

5.6 Flowability versus WFT 100

5.7 Rheological properties versus WFT 101

5.8 Adhesiveness versus WFT 102

5.9 Cube strength versus WFT 103

- viii -

List of Symbols

α, β, γ Symbols Denoting Different Types of Cementitious Materials

φ Solid Concentration

φmax Maximum Solid Concentration

γɺ Shear Rate

ρα Solid Density of Cementitious Material α

ρβ Solid Density of Cementitious Material β

ργ Solid Density of Cementitious Material γ

ρs Solid Density of Fine Aggregate

ρω Density of Water

τ Shear Stress

τ0 Yield Stress

Aα Specific Surface Area of Cementitious Material α

Aβ Specific Surface Area of Cementitious Material β

Aγ Specific Surface Area of Cementitious Material γ

AM Specific Surface Area of Solid Particles in Mortar

As Specific Surface Area of Fine Aggregate

M Mass of Mixture inside Container

Rα Volumetric Ratio of Material α to Total Solid Content

Rβ Volumetric Ratio of Material β to Total Solid Content

Rγ Volumetric Ratio of Material γ to Total Solid Content

Rs Volumetric Ratio of Fine Aggregate to Total Solid Content

V Volume of Container

n Empirical Rheological Coefficient (Non-Dimensional)

u Voids Ratio

uw Water Ratio

uw′ Excess Water Ratio

- ix -

List of Abbreviations

CSF Condensed Silica Fume

FAM Fly Ash Microsphere

HPC High-Performance Concrete

HSC High-Strength Concrete

OPC Ordinary Portland Cement

SCM Supplementary Cementitious Material

SFC Superfine Cement

SP Superplasticizer

W/C ratio Water/Cement Ratio

W/CM ratio Water/Cementitious Materials Ratio

W/S ratio Water/Solid Ratio

WFT Water Film Thickness

- 1 -

CHAPTER 1

INTRODUCTION

1.1 Background

In less than one century, concrete has rapidly gained its popularity and it is

nowadays the most widely used construction material in the world. According to

CEMBUREAU, in 1900, the total world production of cement was about 10

million tonnes; in 1998 it was 1.6 billion tonnes. Glavind and Munch-Petersen

[2002] calculated and reported that the annual consumption of concrete was about

5 billion m3 in 2002 world widely. Almost at the same time, Mehta [2001]

predicted that the demand of concrete would double every decade. Time has

proved Mehta’s prediction and now it is estimated that over 10 billion m3 of

concrete are produced each year. Due to the huge quantity of consumption, its

performance and environmental footprint on the earth are of great importance.

The use of concrete can be traced back to the Roman Empire, when

Roman concrete was made from quicklime, pozzolana and an aggregate of pumice.

After the Empire passed, use of concrete became scarce until the technology was

re-pioneered in the mid-18th century. The past 50 years has witnessed the

significant evolvement in concrete technology. Before the 1960s, concrete was

still a mixture of only cement, water and aggregates. The concrete produced then

was mostly 1:2:4 or 1:1:2 (Portland cement: fine aggregate: coarse aggregate)

nominal mix whose compressive strength was usually between 15 to 25 MPa.

Although the water/cement (W/C) ratio had been known as the controlling factor

governing the strength, the water content was not specified in the common mix

design of concrete. Hence, the poor and varied quality of concrete could be

expected.

The earliest need for concrete was from the aspect of a higher strength.

Lowering the W/C ratio is an effective way but limited by the workability

requirement. There should be sufficient workability for the concrete to achieve

- 2 -

adequate compaction at the low W/C ratio [Neville, 1995]. It was reported that

Parrott [1969] had produced concrete with 28-day cube strengths of 80-110 MPa

using just ordinary Portland cement (OPC), rock aggregate and water. But due to

the adoption of the very low W/C ratios of 0.20-0.30, the concrete mixes only had

a slump smaller than 25 mm. Until the high-range water reducers or known

presently as superplasticizers (SP) was developed almost simultaneously in Japan

and Germany in the 1970s [Hattori, 1978; Meyer, 1979], this situation started a

great change. The SP could significantly improve the workability of concrete or

reduce the water demand so that the W/C ratio could be reduced to achieve higher

strength. In recent years, with the newest generation polycarboxylate-based SP

being developed and applied, the strength or workability performance of concrete

is further improved and now much higher than ever before [ACI Committee 212,

2004; Collepardi and Valente, 2006].

The condensed silica fume (CSF), which is a by-product of semiconductor

industry, was found to be an excellent mineral admixture to effectively enhance

the performance of concrete especially when it was added together with SP. The

CSF is not only a highly reactive pozzolan, but also of high fineness, the particles

of which are about 100 times smaller than the ones in cement. Both of the two

attributes are beneficial to the strength of concrete. The extremely fine particles of

CSF can pack tightly against the surface of the aggregate and fill in the voids

between cement particles thus greatly improve the packing. As a result, the size

and volume of voids near the surface of the aggregate was reduced, the quality of

transition zone was improved so as the bond between the aggregate and cement

paste. Hence, the strength was significantly increased as well as the durability due

to the decreased permeability. With proper addition of SP and CSF, a very high

concrete strength of up to 150 MPa can be produced, being mainly limited by the

aggregate used [Kwan et al., 1995].

The high strength of concrete allows the designer to use smaller columns

and, therefore, reduce the total weight of tall buildings, and hence, decrease the

load on the foundations. Moreover, the horizontal area occupied by the columns

would be reduced so that there would be more usable area, which is of high

economical value in the place where land is scarce, like in Hong Kong. The

- 3 -

development of high-strength concrete (HSC) was very fast after first introduced

in 1990 [Read and MacArthur, 1990]. Later, Grade 100 HSC was used in 1996 in

Hong Kong [MacArthur and Read, 1996]. Besides the pleasant high strength, the

early HSC was considered to have two major problems, high cement content and

brittle failure respectively. The high cement content resulted in large thermal

expansion/contraction at the early age and large dying shrinkage at the later age,

hence, unsatisfactory dimensional stability of the HSC. On the other hand, the

brittleness of HSC would reduce the ductility to a dangerously low level. To avoid

such brittle failure, the concrete strength is suggested not to exceed 80% of the

ceiling strength in the mix design [Kwan, 2003]. To avoid the high cement content

in concrete, other ways to lower the water/cementitious materials (W/CM) ratio,

besides single increasing the cement content, should be established.

After many years of development of HSC, the emphasis has moved from

the very high strength to other properties under different circumstances. The

market has been demanding the concrete that has not only high strength but also

all around high performance in terms of workability, dimensional stability and

durability, which promoted a dramatic advancement of high-performance concrete

(HPC) since 1990s. Now it is not difficult to produce a HPC having a

characteristic cube strength higher than 100 MPa, a workability high enough for

pumping up to higher than 300 m and placing without vibration, a temperature

rise small enough to render cooling during curing unnecessary and a durability

good enough to endure more than one century even for the concrete structures in

marine environment. There is no doubt that concrete technology is one of the most

rapidly advancing areas in civil engineering in the last twenty years.

1.2 Development of High-Performance Concrete

To produce HPC, a low W/CM is still an essence due to the requirements

of strength and durability. However, it is not possible to unlimitedly reduce the

W/CM ratio since the water should be more than sufficient to fill the voids

between the cement grains so that provide the water-solid mixture workability.

- 4 -

Powers [1963] proposed a similar “excess paste theory”, indicating that it is the

excess paste (paste in excess of that needed to fill the voids between aggregate

particles) that gives the mortar or concrete workability. Hence, how low the

W/CM ratio could be reduced is dependent on the workability requirements

whereas the workability of concrete mixes is dependent on the packing density of

solid particles and water content together. In general, the most difficult part in

producing HPC lies on the facts that the various required performance attributes,

like high strength, high workability, high durability and high dimensional stability

and so on, usually impose contradictory requirements on the mix parameters to be

adopted. For example, increasing the strength usually leads to decrease in

workability and vice versa, and increases in both strength and workability might

have to be achieved at the expense of a lower dimensional stability. As a result,

the mix design has remained a difficult task since the very beginning of the

development of HPC.

Supplementary cementitious materials (SCM) were found of great benefits

in producing HPC. Firstly, the fine SCM can increase the packing density so that

improve the flowability of concrete at a low W/CM ratio. Since the fine SCM

blended with cement will fill the voids between the cement grains so that the

water trapped in those voids will be released as excess water (water in excess of

that needed to fill up the voids) to provide better lubrication [Feng et al., 2000;

Kwan, 2000; Yahia et al., 2005]. Secondly, partly replacement of cement by SCM

will improve the dimensional stability at the same time maintain sufficient

strength and workability. The use of fly ash to replace cement in concrete was

recorded about forty year ago in many countries [ACI Committee 226, 1987]. In

2000, the utilization rate of fly ash in Hong Kong is 93.7%, which ranked the

highest in the world according to Malhotra’s study [2000]. Other SCM, including

ground granulated blast slag, rice husk ash, metakaolin, etc., are added more

commonly than before in the production of HPC. Due to they are usually industry

by-products, addition of SCM can not only reduce the costs and improve the

performance of concrete, but also benefits the greening of concrete industry in the

long run.

- 5 -

However the effects of SCM on the fresh state properties of concrete are

still elusive as reflected by the different or even contradictory results reported by

different researchers. In 1986, Kohno and Komatsu [1986] reported that the

addition of CSF would, at same W/CM ratio and SP dosage, decrease the

flowability of a mortar. Later in 1998, Duval and Kadri [1998] yet demonstrated

that the addition of CSF had no adverse effect on concrete workability if the

replacement content was no more than 10% by mass. In 1999, Collins and

Sanjayan [1999] found that the addition of ultrafine fly ash could increase the

flowability but the addition of ultrafine slag would slightly decrease the

flowability. Then, in 2001, Gao et al. [2001] reported that replacement of cement

by superfine slag up to 20% could improve the workability. These studies focused

on the single addition of one kind of SCM whereas situation might be more

complex when a ternary cementitious system is used. Without a clear and

fundamental understanding of the mechanism that how SCM influence concrete’s

rheology, it is never possible to develop a scientific and systematic design method

to use SCM in HPC.

Based on the experimental results obtained during the author’s

postgraduate study, it is found that not only the packing density of the particle

system but also the solid surface area would majorly govern the rheology of

cement paste/mortar/concrete. This is considered to be one of the reasons that the

addition of SCM might or might not improve the flowability, due to the relative

magnitudes of the two opposite effects. Actually these two major governing

parameters could be simply combined into one single parameter called water film

thickness, taken as the average thickness of water films coating the solid particles

and evaluated as the excess water to solid surface area ratio. Besides the water

film thickness, the effects of particle interaction, for example, the ball bearing

effect of some SCM particles would also play important roles in the rheology of

concrete. The afore-mentioned theories of packing of solid particles, water film

thickness, and particle interaction lie on the scope of particuology, which refers to

the science of particle systems. Research on these theories is still preliminary but

in the author’s belief, could transform the traditional concrete technology, which

is more know-how oriented, into modern concrete science, which is more know-

- 6 -

why oriented. The roles of these new concepts in the performances of HPC will be

presented and discussed in details in the later chapters.

1.3 Three-Tier Mix Design Approach

To systematically develop a mix design guideline for HPC, the author’s

research team launched a “three-tier system”, which categorizes the concrete mix

into three tiers, as depicted below:

Cement paste: cementitious materials + water + SP

Mortar: cement paste + aggregate particles smaller than 1.2 mm

Concrete mix: mortar + aggregate particles larger than 1.2 mm

Compared with the conventional two-tier system, which categorizes the

concrete mix into only cement paste and concrete mix, this new developed three-

tier system has one additional tier – the mortar. This categorization comes from

the observation that even when the concrete mix is showing serious signs of

segregation, the aggregate particles smaller than 1.2 mm tend to stay with the

cement paste. Therefore, it is believed that the aggregates with particles smaller

than 1.2 mm behave differently from the ones with particles larger than 1.2 mm.

The coherent mixture of cement paste and aggregate particles smaller than 1.2 mm,

which is the mortar formed, should have great impact on the performance of

concrete mix. But the common definition of fine aggregate is that aggregate with

particles smaller than 4.75 mm, which, from the author’s point of view, should be

further divided. When the aggregates size ranges between 1.2 and 4.75 mm, they

won’t be cohered with the cement paste to form mortar so that they should be

taken into consideration separately with the aggregates smaller than 1.2mm in the

mix design. This finer division into three tiers is also from the hope of developing

a more optimum mix design method. The author’s postgraduate study focuses on

the mortar tier to investigate the effects of different SCM through the concepts of

packing density, water film thickness and particle interaction.

- 7 -

1.4 Development of Green Concrete

With the increasing public concerns about sustainable development, the

concept of green concrete is raised with high frequency. In the author’s point of

view, the HPC itself, which consists of high volume replacing SCM and renders

less total volume of concrete for structural needs and longer service life, is also a

kind of green concrete and the experience of developing HPC would be of high

value in the development of green concrete. In general, green concrete can be

defined as concrete with less cement consumption due to its energy-intensive

production process together with the large amount of carbon emissions it accounts

for. It is widely accepted that producing one ton of cement releases approximately

one ton of carbon dioxide, and the yearly cement production would be responsible

for about 7-8% of global loading of carbon dioxide released the atmosphere

finally [Malhotra, 1999; Mehta, 1999]. Using SCM to replace cement in the

production of HPC would be of great potential to mitigate this problem. There

have been extensive proposals to develop new SCM to minimize the cement

consumption. On the other hand, the conservation of aggregate should be also

concerned in green concrete. Researches on recycled concrete, post-consumer

glass, recycled tires and other recycled materials have been in progress. Moreover,

Mehta [2001] pointed out that by better aggregate grading and by expanding the

use of SCMs and SP, the yearly global mixing water amount, which is 1 trillion

litres by then, can be cut in half. In a word, green concrete will be the major trend

of concrete industry and the use of SCM and the science of particuology will play

important roles in the development of green concrete for future.

1.5 Objectives of Research

The objectives of this research are summarized as follows:

1 To estimate a comprehensive mechanism explaining how the SCM

improve the overall performance of mortar.

2 To investigate the role of packing density, water film thickness and

- 8 -

particle interaction in the fresh and hardened properties of mortar.

3 To study the effects of fly ash microsphere (FAM) on improving the

packing density, rheology and strength of mortar.

4 To study the effects of superfine cement (SFC) on improving the packing

density, rheology and strength of mortar.

5 To study the effects of ternary blending FAM and CSF on improving the


6 To study the effects of ternary blending SFC and CSF on improving the


7 To investigate the combined effects of the particle interaction and water

film thickness on rheology and strength of mortar.

1.6 Organization of Thesis

The organization of this thesis is established as follows:

Chapter 2 presents the experimental program of measurement of packing

density, rheology and strength of mortar. The test methods adopted for these

measurements are stated in detail in this chapter. A new wet packing method is

developed and a new parameter, water film thickness, is proposed and evaluated

from the packing density result.

Chapter 3 investigates the effects of fly ash microsphere (FAM) on

packing density, rheology and strength of mortar through 22 mortar samples with

different FAM contents at various W/CM ratios. The effects of FAM on the fresh

and hardened properties in mortar are illustrated and the concurrent strength-

flowability performance of FAM mortar is examined. The effect of FAM on water

film thickness and the role of water film thickness in various performances of

mortar are discussed in the in-depth analysis.

Chapter 4 extends the study in Chapter 3 to further evaluates the effects of

combined addition of FAM and CSF on packing density, rheology and strength of

- 9 -

mortar through 42 mortar samples with different FAM and CSF contents at

various W/CM ratios. The effects of FAM and CSF on these properties under

varying W/CM ratios are illustrated and the overall strength-flowability of mortar

is examined. The effects of FAM and CSF on water film thickness and the role of

water film thickness on various performances of mortar are discussed with in-

depth analysis. The filling effect, ball bearing effect and CSF-slurry effect are

studied preliminarily in this chapter.

Chapter 5 studies the effect of combined addition of SFC and CSF on

packing density, rheology and strength of mortar through 27 mortar samples with

different SFC and CSF contents at various W/CM ratios. The effects of SFC and

CSF on water film thickness and the role of water film thickness on various

performances of mortar are discussed with in-depth analysis.

- 10 -

CHAPTER 2

EXPERIMENTAL PROGRAM

2.1 Measurement of Packing Density

The packing density of the particle system was measured by a wet packing

method with the solid particles mixed with water so that the solid particles were

under wet condition. Basically, the wet packing method determines the packing

density of the solid particles in a mortar as the maximum solid concentration of

the solid particles that can be achieved at different water/solid (W/S) ratios. This

method usually starts the measurement at a relatively high W/S ratio and reduces

the W/S ratio until the measured solid concentration had increased to a peak value

and started to decrease. Usually six to eight samples would be made to conduct

this measurement. The detailed procedures are summarized as below:

i) Set the W/S ratio at which the test is to be carried out. Weigh the required

amounts of water, cementitious materials, fine aggregate and

superplasticizer and dose each ingredient into a separate container.

ii) Pre-mix the cementitious materials with fine aggregate in dry for 2

minutes to improve the uniformity of the aggregate sample.

iii) Add all the water into the mixing bowl.

iv) Add half of the cementitious materials, fine aggregate and superplasticizer

into the mixing bowl and run the mixer at low speed for 3 minutes.

v) Add the remaining cementitious materials, fine aggregate and

superplasticizer in four times with four equal portions and after each

addition run the mixer for 3 minutes.

vi) Transfer the mixture to a cylindrical container and fill the container in

three to five layers. In each layer, compaction is applied to rule out the

trapped air. After the last layer, remove the excess mortar with a straight

edge. Weigh the amount of mixture in the container.

vii) Repeat steps (i) to (vi) at successively lower W/S ratios until the maximum

solid concentration has been obtained.

- 11 -

From the test results obtained, the solid concentration of the solid particles

in the mortar can be determined as follows. Define M as the mass of the mixture

in the container while V as the volume of the cylindrical container which has been

known. The wet bulk density of the mortar mix is equal to M/V and the solid

concentration φ can be worked out as:

φ = s sw w

M V

u R R R Rα α β β γ γρ ρ ρ ρ ρ+ + + + (2.1)

where ρw is the density of water, ρα, ρβ and ργ are the densities of the cementitious

material α, β and γ, ρs is the solid densities of fine aggregate, uw is the water ratio

(same as the W/S ratio by volume) and Rα, Rβ, Rγ and Rs are the volumetric ratios

of α, β, γ and fine aggregate to the total solid content. The maximum value of φ so

obtained is taken as the packing density of the solid particles in mortar.

2.2 Determination of Water Film Thickness

Based on the packing density result, the voids ratio of the particle system

can be determined as:

u = max

max1

φφ−

(2.2)

where u is the voids ratio (the ratio of the volume of voids in the bulk volume to

the solid volume of the solid particles) and φmax is the maximum solid

concentration of the solid particles. From the voids ratio so determined, the excess

water ratio of the mortar can be evaluated as:

uw′ = uw − u (2.3)

where uw′ is the excess water ratio and uw is the water ratio (same as the W/S ratio

by volume) of the mortar. This excess water ratio has the physical meaning of

- 12 -

being the amount of excess water in the mortar per solid volume of the particles.

Meanwhile, the specific surface area (defined as solid surface area per unit solid

volume) in the mortar AM can be calculated as:

AM = Aα × Rα + Aβ × Rβ + Aγ × Rγ + As × Rs (2.4)

in which Aα, Aβ, Aγ and As are respectively the specific surface areas of

cementitious materials α, β and γ and fine aggregate, while Rα, Rβ, Rγ and Rs are

respectively the volumetric ratios of α, β, γ and fine aggregate to the total solid

volume. With the values of uw′ and AM so obtained, the water film thickness (WFT)

can be calculated as:

WFT = M

w

A

'u (2.5)

2.3 Measurement of Flowability

The flowability of each mortar sample was evaluated in terms of flow

spread and flow rate using the mini slump cone test and mini V-funnel test

respectively. Both of the mini slump cone and mini V-funnel tests for mortar

could be considered as reduced scale versions of the slump and V-funnel tests for

concrete. In this study, the mini slump cone and mini V-funnel tests adopted were

the same as those used by Okamura and Ouchi [2003]. The mortar samples to be

tested are prepared by the same mixing procedure as in the wet packing test. The

dimensions of the slump cone and V-funnel are shown in Figure 2.1(a) and Figure

2.1(b) respectively.

The procedures of the mini slump cone test are detailed as follows:

i) Place the slump cone at the centre of a leveled, flat and smooth steel plate.

- 13 -

ii) Pour the mortar slowly into the slump cone until the slump cone is

completely filled up and trowel flat the mortar’s top surface with a straight

edge.

iii) Lift the slump cone gently and allow the mortar to flow and spread till

stoppage.

iv) Measure the diameters of the mortar patty formed at two perpendicular

directions, calculate the average diameter and take the average diameter

minus the base diameter of the slump cone as the flow spread of the mortar.

The procedures of the mini V-funnel test are detailed as follows:

i) Mount the V-funnel on a stable stand and keep the opening at the bottom

closed.

ii) Pour the mortar into the V-funnel slowly along the inner surface until the

V-funnel is completely filled up. Trowel flat the mortar’s top surface with

a straight edge.

iii) Open the opening of the V-funnel and start timing. Take the time from the

start of the flow to the first appearance of light through the opening as the

flow time of the mortar.

As the flowability of the mortar is inversely proportional to the flow time,

the test result is transferred in terms of the flow rate, which can be calculated as

the volume of the mortar in the V-funnel divided by the flow time.

2.4 Measurement of Rheological properties

The shear vane test was used to evaluate the rheological properties of the

mortar samples. It was carried out using a speed-controlled rheometer equipped

with a shear vane, measuring 20 mm in width and 40 mm in length, and a

cylindrical container, having an inner diameter of 40 mm, as shown in Figure 2.2.

The inner wall of the container was profiled with grooves of which the asperity

- 14 -

was larger than the largest particle in the mortar to minimize slippage of the

mortar during shearing.

At the onset of the test, the shear vane was concentrically inserted into the

mortar sample in the cylindrical container and then set to rotate at controlled

rotation speed, following a shearing sequence which consisted of two shearing

cycles. The first shearing cycle, called pre-shearing cycle, was to ensure all the

samples tested had undergone the same shearing history before measurement. The

second shearing cycle, called data-logging cycle, was the cycle in which actual

measurement was carried out. In each shearing cycle, the rotation speed N

(measured in terms of rpm, i.e. rotation per minute) was increased from 0 to 50

rpm in 75 seconds and then decreased to 0 rpm in another 75 s. In other words, the

corresponding shear rate was increased from 0 to 14 s-1 in 75 s and then decreased

from 14 to 0 s-1 in another 75 s. The shear stress-shear rate curves obtained at

decreasing shear rate, which is generally more consistent and repeatable, was used

for evaluating the rheological properties of the mortar sample.

Due to the mortar is non-Newtonian, it is customary to describe its

rheological properties by either the Bingham model or the Herschel-Bulkley

model. The first one assumes the shear stress-shear rate curve is linear while the

later one assumes the shear stress-shear rate curve follows the power equation.

Through curve fitting using both models, it is found that the experimental results

agrees better with the Herschel-Bulkley model, whose shear stress-shear strain

equation is given by:

nkγττ ɺ+= 0 (2.6)

where τ is shear stress (Pa), γɺ is shear rate (s-1), 0τ is yield stress (Pa), and K

(Pasn) and n (non-dimension) are empirical coefficients. For each mortar sample,

the best-fit curve based on the above equation was obtained by regression analysis

and from the best-fit curve so obtained, the shear stress at a shear rate of 0 s-1 and

the ratio of shear stress to shear rate at a shear rate of 14 s-1 were taken as the yield

stress and the apparent viscosity, respectively.

- 15 -

2.5 Measurement of Adhesiveness

Various tests, for example the probe tack test [Kaci et al. 2011], have been

developed to measure the adhesiveness of cement paste and mortar, but so far

none has become standardized. Herein, a new test, called stone rod adhesion test,

was developed recently by the author’s research team [Li and Kwan, 2011]. The

apparatus, shown in Figure 2.3, consists of a handle with six stone rods vertically

fixed underneath and a container. The stone rods are made of granite, which is a

commonly used rock for coarse aggregate, and each stone rod has a diameter of 10

mm and an exposed length of 110 mm. Before the test, the stone rods were

immersed in water for at least 24 hours and then wiped clean by a piece of dry

cloth so that the stone rods were “saturated and surface dry”. During the test, the

stone rods were immersed into the mortar inside the container with an immersion

depth of 100 mm, as indicated by the mortar surface reaching the 100 mm mark

on the stone rods. The stone rods were left immersed in the mortar for 1 minute

and afterwards pulled out steadily and slowly. The handle holding the stone rods

was then placed on a stand to allow dripping to take place. After several minutes

when no more dripping occurred, the increase in weight of the handle (in other

words, the weight of mortar adhering to the stone rods) was measured and taken

as the adhesiveness of the mortar tested.

2.6 Measurement of Strength

Three 100 mm cubes were made from each mortar sample for strength

measurement. The cubes were made by placing the mortar into a cube mould,

inserting a vibrator into the mortar for compaction and covering the top surface of

the mould with a plastic sheet. After casting, the cubes were stored at a

temperature of 24 ± 2 °C. After one day, the moulds were removed and the cubes

were cured in a water tank controlled at a temperature of 27 ± 2 °C until the age of

28 days for cube compression test.

- 16 -

Figures

(a)

Figure 2.1 (a) Mini slump cone; (b) Mini V-funnel

60 mm

100 mm

70 mm

30 mm

60 mm

240 mm

30 mm

270 mm

(b)

- 17 -

Figure 2.2 Rheometer, shear vane and container

Rheometer

Shear vane

Container

- 18 -

Figure 2.3 Details of apparatus for stone rod adhesion test

30 mm

30 mm

30 mm

Immersion depth = 100 mm

Length = 110 m

Diameter = 10 mm

- 19 -

CHAPTER 3

BINARY CEMENTITIOUS SYSTEM CONTAINING FLY ASH

MICROSPHERE

3.1 Introduction

Supplementary cementitious materials (SCMs) are nowadays widely used

in concrete. The addition of certain SCMs, which are by-products or wastes of

industrial manufacturing processes, such as fly ash and ground granulated

blastfurnace slag, to replace part of the cement is an effective way of reducing the

carbon footprint of our concrete production [Aïtcin, 2000]. If the SCM is finer

than cement, its addition to fill into the voids in cement would increase the

packing density of cementitious materials so as to improve the performance of

concrete [Aïtcin, 1998]. In 1994, by maximizing the packing density of

cementitious materials, de Larrard and Sedran [1994] lowered the

water/cementitious materials (W/CM) ratio to 0.14 and thus achieved a

compressive strength of 236 MPa. Many literatures also reported that concrete

containing SCM often perform better in terms of workability, robustness, strength

and durability [Khatri et al., 1995; Kwan and Ng, 2010; Lothenbach et al., 2011].

Recently, in 2011, Aïtcin [2011] pointed out that with SCMs finer than cement

(hereafter referred to as superfine SCMs) added to fill into the voids of the particle

system, the concrete will bleed less, shrink less and attain higher strength.

Previous studies [Kwan and Wong, 2008b; Lee et al., 2003;

Nanthagopalan et al., 2008] have revealed that the packing density of cementitious

materials is an important factor governing the flowability of cement paste,

especially at low W/CM ratio. With superfine SCMs added to fill into the voids

between cement grains, the packing density of the cementitious materials would

become higher and the volume of voids in the cementitious materials would

become smaller. As a result, the amount of water needed to fill the voids would

- 20 -

decrease and, at the same W/CM ratio, the amount of excess water (water in

excess of that needed to fill the voids) available for forming water films coating

the solid particles to provide lubrication would increase. However, it has been

found that increasing the packing density does not always increase the flowability

and that apart from the packing density and water content, the type and amount of

SCM added also have certain effects.

Collins and Sanjayan [1999] showed that ultrafine fly ash and condensed

silica fume have widely different effects on the workability and strength of

concrete. Claisse et al. [2001] concluded that the addition of silica fume could

impair the flowability of cement paste because of its large specific surface area,

which increases the water demand. Ferraris et al. [2001] postulated that the

increase in packing density and the increase in solid surface area due to the

addition of superfine SCMs have opposite effects on the rheology of cement paste.

To reflect the combined effects of packing density and solid surface area, Kwan

and Wong [2008a] proposed a new mix parameter - water film thickness (WFT),

which is defined as the average thickness of the water films coating the solid

particles and may be determined as the excess water to solid surface area ratio.

This WFT was, in subsequent studies [Fung and Kwan, 2010; Kwan et al., 2010;

Wong and Kwan, 2008a], demonstrated to be a major factor governing the

rheology of cement paste and mortar.

Due to agglomeration caused by inter-particle forces, the packing density

of fine solid particles is not easy to measure [Yu et al., 1997]. The existing dry

packing methods, which measure the packing density under dry condition, have

the major problems that the measured results are sensitive to the amount of

compaction applied [Svarovsky and Board, 1987] and that they cannot take into

account the possible effects of water and chemical admixtures. In 1999, de

Larrard [1999] proposed a wet method which takes the minimum volume of water

required to change the water-solid mixture from a solid lump to a slurry as the

voids volume for the determination of packing density. However, it is often

difficult to judge very precisely the point of transition from solid lump to slurry

because the transition is actually gradual. More recently, the authors’ research

team has developed a wet packing method for direct measurement of the packing

- 21 -

densities of cementitious materials [Wong and Kwan, 2008b], fine aggregate

[Fung et al., 2009] and mortar [Kwan and Fung, 2009]. Basically, this wet

packing method measures the solid concentration of the solid particles at different

water contents and takes the maximum solid concentration as the packing density

of the solid particles. Its accuracy and applicability have been verified by

comparing with theoretical results based on established packing models [Kwan

and Fung, 2009; Wong and Kwan, 2008c].

On the other hand, it is well recognized that the particle size distribution

and size range have great effects on the packing density and rheological

performance of cement paste, mortar and concrete [de Larrard and Sedran, 1999;

Fung et al., 2009; Kwan and Fung, 2009; Kwan and Wong, 2008b; Wang et al.,

1997; Wong and Kwan, 2008c]. Ideally, the particle system should have a wide

range of particle size so that the medium size particles would fill into the voids of

the larger size particles, the fine particles would fill into the voids of the medium

size particles and the superfine particles would fill into the voids of the fine

particles and so on. In theory, such successive filling of voids by a continuous

spectrum of finer and even finer particles can greatly reduce the voids volume and

increase the packing density. Aïtcin [2011] pointed out that there are two gaps in

the particle size distribution of the solid ingredients in high-performance concrete

and suggested to fill up these gaps to further enhance the concrete performance.

Whilst the size gap between the cement and fine aggregate can be filled by adding

limestone fine or ground sand about 50 µm in size, the size gap between the

cement and condensed silica fume has to be filled by a superfine powder several

µm in size.

The author is proposing that fly ash microsphere (FAM), which is a

superfine fly ash collected from the exhaust smoke of coal-fired power station,

may be a suitable material for extending the size range of the cementitious

materials or filling the size gap between the cement and condensed silica fume.

The FAM that the authors have obtained for testing is perfectly spherical in shape

and has an average diameter of about 3 µm. Being superfine, it can fill into the

voids in cement to increase the packing density of the cementitious materials.

- 22 -

Compared with condensed silica fume, FAM has a much smaller specific surface

area and thus can increase the packing density without excessively increasing the

total solid surface area to be coated with water films. Herein, a comprehensive

testing program was carried out to study the effects of FAM on the packing

density and overall performance of mortar and the roles of WFT in the rheology,

adhesiveness and strength of mortar.

3.2 Materials

The cement used was an ordinary Portland cement (OPC) obtained from

the local market in Hong Kong. It was of strength class 52.5N and had been tested

to comply with BS EN 197-1: 2000. The FAM used was imported from China

which according to the supplier had been tested to comply with Chinese National

Standard GB 1596-91. The chemical compositions of OPC and FAM, as provided

by the suppliers, are presented in Table 3.1. The fine aggregate used was a local

crushed granite rock fine with a maximum size of 1.18 mm and a water absorption

of 1.02% by mass. The relative densities of the OPC, FAM, and fine aggregate

had been measured in accordance with BS EN 196-6: 2010 as 3.11, 2.52 and 2.48,

respectively. The particle size distributions of the materials were measured by a

laser diffraction particle size analyser and the results are plotted in Figure 3.1.

Based on these particle size distributions, the specific surface areas of the OPC,

FAM and fine aggregate were calculated as 1.12×106 m2/m3, 3.95×106 m2/m3, and

1.48×105 m2/m3, respectively, and the mean particle sizes of the OPC, FAM and

fine aggregate were calculated as 14.1 µm, 2.6 µm and 508 µm, respectively.

Compared with the angular shape of the OPC and fine aggregate particles, the

FAM particles were found to have perfectly spherical shape, as depicted by the

scanning electronic microscope image in Figure 3.2.

A superplasticizer (SP) was added to each mortar sample to disperse the

cementitious materials and reduce agglomeration. The SP employed was a

commonly used polycarboxylate-based SP with a solid mass content of 20% and a

relative density of 1.03. Since SP is a surface reactant and it is the SP dosage per

- 23 -

solid surface area that actually governs the effectiveness of the SP [Wong and

Kwan, 2008a, Kwan et al., 2012], the SP dosage was determined according to the

total surface area of the solid particles in the mortar. Before setting the SP dosage

to be used, trial cement paste mixing using different SP dosages was carried out

and it was found that the saturation dosage (the dosage beyond which further

addition of the SP yields little further increase in flowability) was 2.6×10-7 kg/m2

of the solid surface area. Hence, the SP dosage in terms of liquid mass of SP per

solid surface area was set as 2.6×10-7 kg/m2 for all the mortar samples. It is

noteworthy that since the FAM has higher fineness, the SP dosage per mass of

cementitious materials was higher at a higher FAM content.

3.3 Experimental Program

To study the effects of FAM, three different FAM contents, namely 0%,

20% and 40%, each expressed as a percentage by volume of the total cementitious

materials, were adopted for the design of the mortar samples. To exclude the

effect of variation in fine aggregate content, the total cementitious materials to

fine aggregate ratio was fixed at 0.75 by volume. The W/CM ratio by volume was

varied from 0.5 to 1.4. In total, 22 mortar samples were produced for testing. The

mix proportions of the mortar samples are summarized in Table 3.2. Each mortar

sample is assigned a designation of M-X-Y, where M denotes mortar, X denotes

the FAM content and Y denotes the W/CM ratio by volume. For reference, the

W/CM ratio by mass of each mortar sample is listed in the second column of

Table 3.2.

The experimental program consisted of two parts. The first part was to

measure the packing density of the solid particles in each mortar sample by a wet

packing method. In the second part, the flow spread, flow rate, yield stress,

apparent viscosity, adhesiveness and cube strength of the mortar samples were

measured. Each mortar sample was produced using a standard mixer by first

adding all the water and SP to the mixer and then adding the solid materials bit by

bit to the mixer while mixing. This mixing procedure has been proved to be more

- 24 -

effective than the conventional mixing procedure of adding all the water and solid

materials to the mixer in one single batch, especially when the water content is

low and/or a superfine material is added [Wong and Kwan, 2008b]. All of the

mixing and testing procedures were carried out in a laboratory maintained at a

temperature of 24 ± 2 ºC.

3.4 Test Methods

Measurement of Packing Density

The three mixes of solid samples in the mortar samples were subjected to

the packing density test. The details of the test procedures can be referred to

section 2.1 in Chapter 2.

Determination of Water Film Thickness

The water film thickness (WFT) of the mortar samples, each determined as

the respective excess water to solid surface area ratio, is calculated. The details of

the calculation procedures can be referred to section 2.2 in Chapter 2.

Measurement of Flowability

Each of the mortar samples was subjected to the mini slump cone test and

mini V-funnel test for evaluation of its flowability in terms of flow spread and

flow rate respectively. The details of the two tests can be referred to section 2.3 in

Chapter 2.

Measurement of Rheological properties

Each of the mortar samples was subjected to the rheometer test for

evaluation of its rheological properties in terms of yield stress and apparent

- 25 -

viscosity respectively. The details of this test can be referred to section 2.4 in

Chapter 2.

Measurement of Adhesiveness

Each of the mortar samples was subjected to the stone rod adhesion test for

evaluation of its adhesiveness. The details of this test can be referred to section

2.5 in Chapter 2.

Measurement of Strength

Each of the mortar samples was subjected to the cube crushing test for

evaluation of its 28-day cube strength. The details of this test can be referred to


3.5 Experimental Results

3.5.1 Packing density and voids ratio

The measured packing densities of the mortar mixes containing different

FAM contents are presented in Figure 3.3. From these results, it can be seen that

with no FAM added, the mortar mix was measured to have a packing density of

0.735. With 20% FAM added, the packing density increased to 0.768 while with

40% FAM added, the packing density further increased to 0.797. In other words,

the addition of 20% FAM and 40% FAM had increased the packing density by

4.5% and 8.4%, respectively. This demonstrates that the very fine FAM particles

could fill into the voids between the cement grains to effectively increase the

packing density. Although the increase in packing density (less than 9%) appears

small, the actual reduction in voids ratio is quite substantial. For better illustration,

the voids ratios of the mortar mixes are calculated by Equation (2.2) and plotted

also in Figure 3.3. It can be seen that with 20% FAM added, the voids ratio was

decreased from 0.361 to 0.302 and with 40% FAM added, the voids ratio was

- 26 -

further decreased to 0.255. In other words, the addition of 20% FAM and 40%

FAM had decreased the voids ratio by 16.3% and 29.4%, respectively. Such

reduction in voids ratio would substantially decrease the amount of water needed

to fill the voids and at the same water content, increase the amount of excess

water available for forming water films.

3.5.2 Water film thickness

The WFT of the mortar mixes, each determined as the respective excess

water to solid surface area ratio as per Equation (2.5), are presented in Figure 3.4.

For the mortar samples tested, the WFT ranged from -0.182 µm to 0.454 µm. It

should be noted that a negative WFT value indicates that the amount of water in

the mortar was not sufficient to fill the voids between the solid particles, leading

to the entrapment of air in the mortar. It should also be noted that since the WFT

is a linear function of the W/S ratio by volume (same as the water ratio) as

depicted by Equation (2.3), all the WFT - W/S ratio curves are straight lines.

From Figure 3.4, it can be seen that at W/S ratio ≤ 0.60, the addition of up

to 40% FAM significantly increased the WFT. The increase in WFT was

generally larger at lower W/S ratio and smaller at higher W/S ratio. For instance,

at a W/S ratio of 0.26, the addition of 40% FAM increased the WFT from -0.182

µm to 0.004 µm (increased the WFT by 0.186 µm) while at a W/S ratio of 0.60,

the addition of 40% FAM increased the WFT from 0.429 µm to 0.454 µm

(increased the WFT by only 0.025 µm). This is because the WFT is governed by

both the excess water ratio and the total solid surface area. The proportional

increase in excess water ratio due to the addition of a superfine SCM is larger at

lower W/S ratio but smaller at higher W/S ratio whereas the proportional increase

in total solid surface area is independent of the W/S ratio. At lower W/S ratio, the

proportional increase in excess water ratio is much larger than the proportional

increase in total solid surface area and thus the increase in WFT is very substantial

whereas at higher W/S ratio, the proportional increase in excess water ratio is only

slightly larger than the proportional increase in total solid surface area and thus

the increase in WFT is just marginal.

- 27 -

Overall, it is evident that the addition of FAM has significant effects on the

packing density, solid surface area and WFT. The effects of FAM on the packing

density and solid surface area are not dependent on the W/S ratio but the effects of

FAM on the WFT are dependent on the W/S ratio. Hence, the net effects of FAM

on the properties of mortar could vary with the water content and the effects of

FAM should be evaluated in conjunction with the water content. Perhaps more

importantly, the addition of up to 40% FAM would allow the W/CM ratio to be

reduced to 0.60 by volume (or 0.20 by mass) without causing the WFT to become

negative (a negative WFT is generally not acceptable). Hence, with FAM added to

increase the packing density, the W/CM ratio may be reduced to improve the

strength and durability of the mortar/concrete produced.

3.5.3 Flow spread and flow rate

The flow spread and flow rate results are tabulated in the second and third

columns of Table 3.3 and plotted against the WFT in Figure 3.5. From the curves

plotted, it is obvious that both the flow spread and flow rate increased with the

WFT. This observed phenomenon was expected, as an increase in WFT should

always improve the flowability of mortar. When the WFT was close to zero, the

flow spread and flow rate were also close to zero due to the lack of excess water

to provide lubrication. It is noteworthy that when the WFT increased, the flow

spread increased with the WFT at a gradually decreasing rate but the flow rate

increased with the WFT at a gradually increasing rate. Hence, the WFT has

different effects on the flow spread and flow rate, albeit an increase in WFT

always increases both the flow spread and flow rate.

Comparing the flow spread - WFT curves and the flow rate - WFT curves

for different FAM contents, it can be seen that at the same WFT, both the flow

spread and flow rate increased with the FAM contents. In other words, even at the

same WFT, the addition of FAM would significantly increase the flowability.

Since the effects of FAM on the packing density and solid surface area have

already been allowed for in the calculation of the WFT, there must be some other

effects of FAM not reflected by the WFT so calculated. It is postulated herein that

- 28 -

the increase in flowability at same WFT due to the addition of FAM may be

attributed to the ball bearing effect of the FAM particles. The FAM particles are

perfectly spherical in shape and finer than the cement grains. Apart from filling

into the voids between cement grains, some of the FAM particles may be located

within the narrow gaps between cement grains thus avoiding direct contact

between the angular cement grains. Being spherical in shape, the FAM particles

can roll quite easily and thereby act like ball bearings to reduce the inter-particle

friction between the cement grains.

3.5.4 Yield stress and apparent viscosity

The yield stress and apparent viscosity results of the mortar samples are

tabulated in the fourth and fifth columns of Table 3.3 and plotted against the WFT

in Figure 3.6. When the WFT was negative or close to zero, the torque required to

shear the mortar sample had occasionally exceeded the capacity of the rheometer

causing the yield stress and apparent viscosity results of some mixes (each shown

in Table 3.3 by a hyphen) to be undetermined. From the curves plotted, it can be

seen that both these two rheological properties gradually decreased as the WFT

increased. This agrees with the general observation that increasing the WFT of a

mortar would reduce the yield stress and apparent viscosity. Comparing the yield

stress - WFT curves and the apparent viscosity - WFT curves for different FAM

contents, it can also be seen that at the same WFT, both the yield stress and

apparent viscosity decreased as the FAM content increased. Such decreases in

yield stress and apparent viscosity as the FAM content increased may be

attributed to the ball bearing effect of the FAM particles, which has been

explained in the previous section.

3.5.5 Correlation between flowability and rheological properties

It is noteworthy that the effects of FAM on the yield stress and apparent

viscosity are similar to the effects of FAM on the flow spread and flow rate. A

possible reason is that the flow spread and flow rate are closely related to the yield

stress and apparent viscosity, respectively, and thus any effects of FAM on the

yield stress and apparent viscosity should also be reflected in the flow spread and

- 29 -

flow rate. To verify this expectation, the flow spread and flow rate are

respectively plotted against the yield stress and apparent viscosity in Figure 3.7.

The flow spread - yield stress curves show that the flow spread - yield

stress relation is not unique. It is dependent to some extent on the FAM content, as

indicated by the shifting downwards of the flow spread - yield stress curve with

increasing FAM content. In other words, the flow spread is dependent not only on

the yield stress but also on some other factors. During the mini slump cone tests, it

was observed that at higher FAM content, the mortar was generally more cohesive

(thicker and stickier). The higher cohesiveness at higher FAM content might have

caused the mortar to become more reluctant to flow horizontally, thus

counteracting the increase in flow spread due to the ball bearing effect of the

FAM particles. As a result, the increase in flow spread is less significant than the

decrease in yield stress and that is why the flow spread - yield stress curve shifts

downwards with increasing FAM content.

On the other hand, the flow rate - apparent viscosity curves show a clear

and unique relation. The flow rate - apparent viscosity curves for different FAM

contents overlap with each other such that a single curve may be derived to fit all

the test results, regardless of the FAM content. Regression analysis of the

correlation between the flow rate and the apparent viscosity has yielded a best-fit

curve with an R2 value of 0.944, as shown in Figure 3.7. The strong correlation

suggests that the flow rate is governed solely by the apparent viscosity. Hence, the

flow rate, which can be measured more easily by the mini V-funnel test, may be

taken as an alternative measure of the apparent viscosity.

3.5.6 Adhesiveness

The stone rod adhesion test results of the mortar samples are tabulated in

the sixth column of Table 3.3 and plotted against the WFT in Figure 3.8.

Generally, when the WFT was negative, the mortar appeared to be rather dry and

the adhesiveness was equal to or very close to zero. As the WFT increased, the

mortar became slightly wetter and the adhesiveness increased dramatically to a

certain maximum value. Then, as the WFT further increased, the mortar became

- 30 -

quite wet and the adhesiveness decreased. It may thus be said that the

adhesiveness is highest when the mortar is neither too dry nor too wet, as

indicated by its WFT, taken as a measure of wetness, falling within the narrow

range of 0.05 to 0.15 µm.

Comparing the adhesiveness curves at different FAM contents, it is

apparent that the maximum adhesiveness is generally higher at higher FAM

content. In other words, the addition of FAM would, at the right wetness,

significantly increase the adhesiveness of the mortar. One probable reason is that

the FAM particles, which fill into the voids between the cement grains, would

increase the number of contact points between the solid particles and thus increase

the inter-particle attractive forces between the solid particles and the adhesive

force between the mortar and the stone rods. Moreover, the optimum WFT for

maximum adhesiveness is generally lower at higher FAM content. Such observed

effects of FAM on adhesiveness provide an important guideline for the design of

high-build mortar for rendering and concrete repair, and the design of the mortar

portion of high segregation stability concrete such as tremie concrete and self-

consolidating concrete.

3.5.7 Cube strength

The 28-day cube strength results are tabulated in the last column of Table

3.3 and plotted against the W/CM ratio by mass and the WFT in Figure 3.9. Each

cube strength result presented is the average of the three cubes casted and tested at

the same time. Generally, the cube strength varied with the W/CM ratio in such a

way that as the W/CM ratio decreased starting from a relatively high value, the

strength increased until a certain peak value was reached and then as the W/CM

ratio further decreased, the strength started to decrease. This is because when the

W/CM ratio decreased while the water was still sufficient to fill the voids between

the solid particles, the water space to be filled with gel products would decrease

and the gel/space ratio would increase, thus causing the strength to increase.

However, when the W/CM ratio decreased to such level that the water was no

longer sufficient to fill the voids, air would be entrapped inside the voids, thus

causing the strength to decrease.

- 31 -

In this particular case, with no FAM added, the maximum cube strength

attained was 96.2 MPa. With 20% FAM added, the maximum cube strength was

increased to 116.2 MPa and with 40% FAM added, the maximum cube strength

was further increased to 121.9 MPa. These results show that the addition of FAM

has great effects on the strength of mortar. It should, however, be noted that at the

same W/CM ratio, the addition of FAM did not always increase the cube strength.

As can be seen from the upper graph in Figure 3.9, at W/CM ratio by mass higher

than 0.30, the cube strength curve for 20% FAM is higher than that for 0% FAM

but the cube strength curve for 40% FAM is lower than that for 20% FAM. Hence,

the addition of FAM to replace part of cement without changing the W/CM ratio

does not necessarily increase the strength. The higher maximum strength at higher

FAM content was actually achieved at lower optimum W/CM ratio. It was the

lowering of the W/CM ratio without causing the entrapment of air made possible

by the increase in packing density due to the addition of FAM that increased the

maximum strength.

From the cube strength - WFT curves plotted in Figure 3.9, it can be seen

that the WFT has certain distinct effects on the strength. When the WFT was

negative, the water was not sufficient to fill the voids and a certain amount of air

was entrapped in the voids causing the strength to be adversely affected. Under

this situation, the strength decreased as the WFT decreased. When the WFT was

positive, the water was more than sufficient to fill the voids. Under this situation,

the strength decreased as the WFT increased because of the corresponding

increase in the W/CM ratio. Hence, the maximum strength was achieved at a WFT

very close to zero, in which case, the water was just sufficient to fill the voids. As

a general rule, therefore, to achieve the highest strength possible, the packing

density should be maximized and a positive but very small WFT should be

adopted in the mix design.

- 32 -

3.6 Concurrent strength-flowability performance

The concurrent strength - flowability performance that can be achieved at

different FAM contents is illustrated by plotting the 28-day cube strength against

the flow spread and flow rate in Figure 3.10. Comparing the performance curves

for different FAM contents, it can be seen that with FAM added, the performance

curves are always shifted upwards and to the right. Since shifting of the curve

upwards would lead to a higher strength at the same flowability and shifting of the

curve to the right would lead to a higher flowability at the same strength, this

indicates that the addition of FAM can increase the strength at the same

flowability, increase the flowability at the same strength, or increase both the

strength and flowability at the same time. Hence, FAM is an excellent SCM for

improving the strength and flowability.

3.7 Conclusions

A number of mortar samples with different FAM contents and different

W/CM ratios were produced for packing density, flowability, rheology,

adhesiveness and strength measurements. On the whole, it was found that the

addition of FAM, which is finer than OPC, could significantly increase the

packing density and WFT of the mortar mix. The increase in WFT was generally

larger at lower water content and smaller at higher water content. At the same

W/CM ratio, this would improve the flowability, or at the same flowability

requirement, this would allow the W/CM ratio to be lowered to increase the

strength.

Correlation of the flow spread, flow rate, yield stress, apparent viscosity

and adhesiveness to the WFT revealed that the WFT is an important factor

governing the fresh properties of mortar. Generally, the flow spread and flow rate

would increase and the yield stress and apparent viscosity would decrease as the

WFT increases. Even at the same WFT, the FAM content has certain beneficial

effects on these properties. Basically, a higher FAM content would at the same

- 33 -

WFT lead to larger flow spread, higher flow rate, lower yield stress and smaller

apparent viscosity. Such effects may be attributed to the ball bearing effect of the

spherical FAM particles, which reduces the inter-particle friction between the

larger size and angular particles. Furthermore, correlation between the flow spread

and yield stress indicated that they are closely related but their relation is

dependent to some extent on the FAM content. Lastly, correlation between the

flow rate and the apparent viscosity revealed that they are uniquely related with an

R2 value of 0.944.

On the other hand, the adhesiveness results showed that the adhesiveness

is highest when the mortar is neither too dry nor too wet, as indicated by its WFT,

taken as a measure of wetness, falling within 0.05 to 0.15 µm. The optimum WFT

for maximum adhesiveness is generally lower at higher FAM content and the

maximum adhesiveness is generally higher at higher FAM content. In other words,

the addition of FAM would, at the right wetness, significant increase the

adhesiveness.

Finally, the strength results revealed that the strength is highest at an

optimum W/CM ratio dependent on the FAM content. At higher FAM content, the

optimum W/CM ratio for maximum strength is lower and the maximum strength

is higher. Plotting the strength against the WFT, it can be seen that the maximum

strength generally occurs at a positive but very small WFT, in which case the

water is just sufficient to fill the voids. Summarising, the effects of FAM on

strength are as follows. At the same W/CM ratio, the addition of FAM would only

marginally increase the strength. However, the addition of FAM would increase

the packing density and thus allow the W/CM ratio to be lowered without causing

the WFT to become negative to increase the strength. It is the lowering of the

W/CM ratio made possible by the increase in packing density that produces the

higher maximum strength at higher FAM content. Furthermore, plotting the

strength against the flowability for different FAM contents, it is evident that the

addition of FAM can increase the strength at same flowability, increase the

flowability at same strength or increase both the strength and flowability at the

same time.

- 34 -

Tables

Table 3.1 Chemical compositions of OPC and FAM

Chemical Proportions (%)

OPC FAM

Calcium oxide (CaO) 65.7 4.8

Silicon dioxide (SiO2) 21.8 56.5

Aluminum oxide (Al2O3) 5.7 26.5

Iron oxide (Fe2O3) 3.6 5.3

Magnesium oxide (MgO) 2.2 1.3

Sulfuric anhydride (SO3) 1.3 0.7

Sodium oxide equivalent (Na2O)eq 0.4 3.6

- 35 -

Table 3.2 Mix proportions of the mortar samples

Mix no. W/CM ratio by

mass

Dosage of each ingredient in the mortar (kg/m3)

OPC FAM Fine

aggregate Water SP

M-0-0.6 0.193 1061 0 1146 202 11.5

M-0-0.7 0.225 1026 0 1108 229 11.2

M-0-0.8 0.257 993 0 1073 253 10.8

M-0-0.9 0.289 963 0 1040 276 10.5

M-0-1.0 0.321 934 0 1009 298 10.2

M-0-1.2 0.386 881 0 952 338 9.6

M-0-1.4 0.450 834 0 901 373 9.1

M-20-0.6 0.200 849 172 1146 198 16.5

M-20-0.7 0.234 821 166 1108 225 16.0

M-20-0.8 0.267 795 161 1073 250 15.5

M-20-0.9 0.301 770 156 1040 273 15.0

M-20-1.0 0.334 747 151 1009 295 14.5

M-20-1.2 0.401 705 143 952 334 13.7

M-20-1.4 0.468 667 135 901 370 13.0

M-40-0.5 0.174 659 356 1187 166 22.2

M-40-0.6 0.209 637 344 1146 194 21.5

M-40-0.7 0.244 616 332 1108 221 20.8

M-40-0.8 0.278 596 322 1073 246 20.1

M-40-0.9 0.313 578 312 1040 269 19.5

M-40-1.0 0.348 560 302 1009 291 18.9

M-40-1.2 0.417 529 285 952 331 17.8

M-40-1.4 0.487 500 270 901 367 16.9

- 36 -

Table 3.3 Flowability, rheological properties, adhesiveness and strength

Mix no. Flow

spread (mm)

Flow rate

(ml/s)

Yield stress (Pa)

Apparent viscosity

(Pas)

Adhesive- ness (g)

28-day cube

strength (MPa)

M-0-0.6 0.0 0.0 - - 0.0 8.7

M-0-0.7 0.0 0.0 - - 0.1 57.5

M-0-0.8 0.0 0.0 - - 0.2 95.0

M-0-0.9 5.5 0.0 - - 2.0 96.2

M-0-1.0 92.0 23.0 47.2 19.5 39.0 89.1

M-0-1.2 182.0 192.2 20.0 6.9 15.7 71.5

M-0-1.4 210.0 405.0 6.0 3.1 7.1 61.0

M-20-0.6 0.0 0.0 - - 0.1 48.2

M-20-0.7 0.0 0.0 - - 0.5 116.2

M-20-0.8 57.5 0.0 - - 23.4 109.7

M-20-0.9 111.5 67.5 32.0 14.0 48.2 107.3

M-20-1.0 151.0 151.2 19.9 8.8 39.5 91.3

M-20-1.2 210.0 378.0 7.0 3.6 19.0 81.9

M-20-1.4 225.0 691.5 3.5 1.7 14.1 66.6

M-40-0.5 0.0 0.0 - - 0.1 117.3

M-40-0.6 0.0 0.0 - - 0.8 121.9

M-40-0.7 62.5 59.7 41.0 22.3 75.2 109.0

M-40-0.8 121.5 143.5 20.4 11.1 43.3 100.9

M-40-0.9 153.5 241.3 14.1 8.1 28 93.0

M-40-1.0 191.0 354.4 9.6 5.0 19.9 84.3

M-40-1.2 238.5 590.6 3.9 1.9 12.9 67.7

M-40-1.4 247.5 859.1 3.2 1.2 10.3 55.4

- 37 -

Figures

0

20

40

60

80

100

0.1 1 10 100 1000 10000

Particle size (µm)

Per

cent

age

pass

ing

(%

) .

Figure 3.1 Particle size distributions of FAM, OPC and fine aggregate

OPC FAM fine aggregate

- 38 -

10000× magnification

Figure 3.2 Scanning electron microscope image of FAM

- 39 -

0.70

0.72

0.74

0.76

0.78

0.80

0.82

0 20 40

FAM content (%)

Pac

king

den

sity

0.20

0.25

0.30

0.35

0.40

0 20 40

FAM content (%)

Voi

ds r

atio

Figure 3.3 Variations of packing density and voids ratio with FAM content

- 40 -

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

0.20 0.30 0.40 0.50 0.60 0.70

W/S ratio by volume

WF

T (

µm)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.4 Variation of WFT with W/S ratio at different FAM contents

- 41 -

0

50

100

150

200

250

300

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Flo

w s

prea

d (m

m)

FAM= 0%

FAM=20%

FAM=40%

0

200

400

600

800

1000

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Flo

w r

ate

(ml/

s)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.5 Flowability versus WFT

- 42 -

0

10

20

30

40

50

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Yie

ld s

tres

s (P

a)FAM= 0%

FAM=20%

FAM=40%

0

5

10

15

20

25

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

App

aren

t vi

scos

ity

(Pas

)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.6 Rheological properties versus WFT

- 43 -

0

50

100

150

200

250

300

0 10 20 30 40 50

Yield stress (Pa)

Flo

w s

prea

d (m

m)

FAM= 0%

FAM=20%

FAM=40%

0

200

400

600

800

1000

1200

0 5 10 15 20 25

Apparent viscosity (Pas)

Flo

w r

ate

(ml/

s)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.7 Correlation between flowability and rheological properties

y=1180x-1.02

R2=0.944

- 44 -

0

20

40

60

80

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Adh

esiv

enes

s (g

)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.8 Adhesiveness versus WFT

- 45 -

0

20

40

60

80

100

120

140

0.15 0.25 0.35 0.45

W/CM ratio by mass

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%

FAM=20%

FAM=40%

0

20

40

60

80

100

120

140

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.9 Cube strength versus W/CM ratio and WFT

- 46 -

0

20

40

60

80

100

120

140

0 50 100 150 200 250 300

Flow spread (mm)

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%

FAM=20%

FAM=40%

0

20

40

60

80

100

120

140

0 200 400 600 800 1000

Flow rate (ml/s)

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%

FAM=20%

FAM=40%

Figure 3.10 Concurrent strength and flowability performance

- 47 -

CHAPTER 4

TERNARY CEMENTITIOUS SYSTEM CONTAINING FLY

ASH MICROSPHERE AND CONDENSED SILICA FUME

4.1 Introduction

High-performance concrete (HPC), with high performance in both fresh

and hardened states, has nowadays gained popularity and is considered the future

of our concrete industry [Aïtcin, 2000]. To produce HPC, it is essential to lower

the water/cementitious materials (W/CM) ratio; as suggested by Neville [1995],

“what makes the concrete a high performance one is a very low water/cement

ratio: always below 0.35, often around 0.25, and occasionally even 0.20”. Such

lowering of the W/CM ratio is made possible by the advent of superplasticizer

(SP), which provides good workability even at very low W/CM ratios. However,

since the water added must be more than sufficient to fill the voids in the bulk

volume of the cementitious materials [Powers, 1968], there is a limit to which the

W/CM ratio can be lowered, no matter how effective the SP is. This limit is not a

constant, but is dependent on the packing density of the cementitious materials [de

Larrard and Sedran, 1994; Sedran et al., 1996], which determines the volume of

voids to be filled with water.

The addition of fine supplementary cementitious materials (SCM) to fill

into the voids between cement grains is an effective way of increasing the packing

density and reducing the volume of voids to be filled with water. With fine SCM

particles filled into the voids, some of the water entrapped therein can be freed as

excess water (the water in excess of that needed to fill the voids) to form water

films coating the solid particles to provide lubrication [Kwan and Wong, 2008b].

This filling effect of fine SCM can produce a denser and more uniform mixture to

improve the strength and durability of concrete [Isaia et al., 2003]. Moreover, the

pozzolanic reaction of fine SCM can produce further gel products to improve the

- 48 -

microstructure [Lothenbach et al., 2011]. In fact, concrete produced with fine

SCMs added is often found to perform better in terms of workability, strength and

durability [Khatri et al., 1995; Papadakis, 2000; Kwan and Ng, 2010]. Various

kinds of SCM, such as fly ash, ground granulated blast furnace slag, silica fume,

metakaolin and rice husk ash, etc, have emerged. Most of the SCMs are industrial

by-products and their use can help to reduce the cement consumption and carbon

footprint of our concrete production [Glavind and Munch-Petersen, 2002; Mehta,

2001; Meyer, 2009; Zheng et al., 2009]. Hence, the addition of SCM has become

a common practice.

The challenge in the application of SCM lies in the mix design. Due to the

additional variable of SCM content, a much larger number of trial batches are

needed to obtain the optimum mixture proportion than that needed for concrete

with no SCM added [Bharatkumar et al., 2001]. A scientific and universally

applicable mix design method is desperately needed but is still lacking [Domone

and Soutsos, 1994; Alves et al., 2004]. Moreover, the effects of SCM on the fresh

properties of concrete are not easy to predict because of their dependence on the

particle size distribution and shape of the SCM. For example, Kohno and

Komatsu [1986] reported that the addition of condensed silica fume (CSF) would

impair the flowability of mortar, while Duval and Kadri [1998] demonstrated that

the addition of CSF up to 10% by mass has no adverse effect on the workability of

concrete. More recently, Kwan and Fung [2011] showed that the addition of CSF

would increase the packing density and thus improve the flowability of mortar.

These contradictory results on the effects of SCM have made the mix design of

HPC containing SCMs a very difficult task.

Although the addition of fine SCM to fill into the voids between cement

grains would no doubt increase the packing density and thus the amount of excess

water available for forming water films, it would at the same time increase the

solid surface area to be coated with water films [Claisse et al., 2001]. Ferraris et al.

[2001] suggested that the increases in packing density and solid surface area have

opposite effects on the rheology of cement paste. Whilst the increase in packing

density would make available more excess water for forming water films, the

increase in solid surface area would thin down the thickness of water films formed.

- 49 -

Therefore, the net effect of adding fine SCM is dependent on the relative

magnitudes of the increase in packing density and the increase in solid surface

area.

To combine these two effects, Kwan and his research team advocated to

use the water film thickness (WFT), the average thickness of water films coating

the solid particles, as the controlling parameter in the mix design of HPC [Kwan

et al., 2010; Wong and Kwan, 2008a]. They have also developed a wet packing

method for direct measurement of the packing density of solid particles in cement

paste and mortar [Wong and Kwan, 2008b; Kwan and Fung, 2009]. From the

packing density so measured, the excess water content may be calculated as the

water content minus the voids volume and the WFT may be determined as the

excess water to solid surface area ratio. The WFT has been found to be the single

most important mix parameter governing the fresh properties of cement paste and

mortar [Fung and Kwan, 2010; Kwan et al., 2010; Kwan and Wong, 2008a].

As mentioned before, the addition of CSF would increase the packing

density and thus release more excess water for forming water films. However, the

large increase in solid surface area due to the high fineness of CSF would thin

down the WFT. Hence, despite increase in packing density, the addition of CSF

may or may not increase the WFT. An intermediate sized SCM that is finer than

cement so as to fill into the voids between cement grains to increase the packing

density but coarser than CSF so as to avoid large increase in the solid surface area

may be more effective in increasing the WFT. Herein, it is proposed that fly ash

microsphere (FAM), which is a superfine fly ash captured from the exhaust smoke

of coal-fired power stations, may be a suitable SCM for such usage. In fact, the

combined use of FAM and CSF, or ternary blending of cement with both FAM

and CSF, may be even better because successive filling of the voids between

cement grains by first the FAM particles and then the CSF particles should further

increase the packing density.

Multiply blending of cement with two or more cementitious materials of

different fineness or pozzolanic reactivity to harvest the synergic effect of the

various cementitious materials is not new [Mehta and Gjørv, 1982; Erdem and

- 50 -

Kırca, 2008]. In this regard, Aïtcin [2011] has pointed out that there is a size gap

between cement and CSF and that this size gap should be closed by adding an

intermediate sized SCM several micrometres in size. The author fully concurs

with this statement and is proposing to ternary blend cement with both FAM and

CSF so as to achieve a better particle size distribution. In this research, such

ternary blending was studied by testing mortar mixes with different amounts of

FAM, CSF and water added. The packing density, flowability, rheology,

adhesiveness and strength of the mortar mixes were measured and the effects of

FAM content, CSF content and WFT on the fresh and hardened properties of

mortar were investigated. It will be seen that the WFT remains a governing factor

in the properties of such kind of mortar and that ternary blending with the size gap

closed is really superior to binary blending.

4.2 Materials

Three types of cementitious materials, namely, ordinary Portland cement

(OPC), FAM and CSF, were adopted in this study. The OPC was of strength class

52.5N obtained from the local market in Hong Kong whereas the FAM and CSF

were imported from Mainland China and Europe, respectively. The OPC, FAM

and CSF had been tested to comply with British Standard BS EN 197-1: 2000,

Chinese Standard GB 1596-91 and American Standard ASTM C 1240-03,

respectively. The fine aggregate adopted was a local crushed granite rock fine

with a maximum size of 1.18 mm and a water absorption of 1.02% by mass. The

solid densities of the OPC, FAM, CSF and fine aggregate had been measured in

accordance with BS EN 196-6: 2010 as 3112 kg/m3, 2520 kg/m3, 2196 kg/m3 and

2537 kg/m3, respectively. Their particle size distributions were measured by a

laser diffraction particle size analyzer and the results obtained are plotted in

Figure 4.1. Using the method proposed by Hunger and Brouwers [2009], the

specific surface areas of the OPC, FAM, CSF and fine aggregate were calculated

from the particle size distribution results as 1.12×106 m2/m3, 3.95×106 m2/m3,

13.3×106 m2/m3 and 0.148×106 m2/m3, respectively. Unlike the cement particles

which are angular in shape, the FAM and the CSF particles are spherical in shape.

- 51 -

The SP added was a polycarboxylate type supplied in the form of an

aqueous solution with a solid mass content of 20% and a relative density of 1.03.

Since SP is a surface reactant and it is the SP dosage per solid surface area that

governs its effectiveness, the SP dosage was expressed in terms of the liquid mass

of SP per solid surface area of the solid particles in mortar [Kwan et al., in press].

Before setting the SP dosage to be used, trial cement paste mixing with various SP

dosages was carried out and it was found that the saturation dosage of the SP (the

dosage beyond which further addition of the SP yields little further increase in

flowability) was 2.6×10-7 kg/m2. Hence, the SP dosage in terms of liquid mass of

SP per solid surface area of cementitious materials was set constant as 2.6×10-7

kg/m2 for all the mortar samples. It is noteworthy that since the FAM and CSF

have higher fineness than the OPC, the SP dosage per mass of cementitious

materials was higher at higher FAM and/or CSF contents.


The experimental program consisted of three parts. The first part was to

measure the packing densities of the mortar samples having different FAM and

CSF contents in order to study the effects of FAM and CSF on the packing

density. The second part was to measure the flow spread, flow rate, yield stress,

apparent viscosity and adhesiveness of the mortar samples produced with different

FAM, CSF and water contents. The last part was to measure the 28-day cube

strength of the mortar samples. The WFT of each mortar sample was determined

from the packing density results obtained in the first part and the W/CM ratio of

the mortar. Then the testing results obtained in the second and third parts were

correlated to the WFT to study the roles of WFT in mortar with ternary blended

cementitious materials.

In this study, the FAM and CSF contents were each expressed as a volume

percentage of the total cementitious materials because the packing density is

governed by volume ratios rather than by mass ratios. Three FAM contents,

namely, 0%, 20% and 40%, and two CSF contents, namely, 0% and 10%, were

- 52 -

adopted for the design of the mortar samples. The W/CM ratio was varied from

0.4 to 1.4 by volume. The detailed mix proportions of the mortar samples are

tabulated in Table 4.1. In the first column, the mix numbers were given in the

format of M-X-Y-Z, where M denotes mortar, X and Y denote the FAM and CSF

contents, respectively, and Z denotes the W/CM ratio by volume. The

corresponding W/CM ratios by mass were listed in the second column of Table

4.1 for easy reference. In total, 42 mortar samples were produced for testing.

Each mortar sample was prepared using a standard mixer by first adding

all the water to the mixer and then adding the solid ingredients and SP bit by bit

into the mixer while mixing. This method has been found to be more effective

than the conventional mixing method of adding all the water and solid ingredients

to the mixer in one single batch, especially when the W/CM ratio is low and/or

ultrafine materials such as CSF are added [Wong and Kwan, 2008a]. All of the

mixing and testing procedures were carried out in a laboratory at a controlled

temperature of 24 ± 2 ºC.

4.4 Test Methods

Measurement of packing density

The six mixes of solid samples in the mortar samples were subjected to the

packing density test. The details of the test procedures can be referred to section

2.1 in Chapter 2.

Determination of WFT





- 53 -




Chapter 2.





Chapter 2.




2.5 in Chapter 2.






4.5.1 Packing density and voids ratio

The packing density and voids ratio results are tabulated in the ninth

column of Table 4.1. With only OPC and no FAM or CSF added, the mortar mix

was measured to have a packing density of 0.735. With OPC blended with FAM,

the packing density was increased to 0.768 at 20% FAM content and to 0.797 at

- 54 -

40% FAM content. With OPC blended with CSF, the packing density was

increased to 0.800 at 10% CSF content. This demonstrates that the addition of

either FAM or CSF can effectively improve the packing density of the solid

particles in mortar. Relatively, the CSF is more effective than the FAM because it

is finer and can fill into the voids between cement grains more readily without

loosening the packing of the cement grains.

When the FAM and CSF were added together, the packing density was

further improved. With FAM already added, the addition of 10% CSF increased

the packing density to 0.821 at 20% FAM content and to 0.839 at 40% FAM

content. On the other hand, with 10% CSF already added, the addition of 20% and

40% FAM increased the packing density to 0.821 and 0.839, respectively. It may

therefore be concluded that ternary blending with FAM and CSF is more effective

than binary blending with either FAM or CSF in improving the packing density.

This is due to successive filling of the voids between the cement grains by the

FAM and CSF particles.

By ternary blending with both FAM and CSF, the packing density was

increased from 0.735 to 0.839 by 14.1%, which at first sight does not appear to be

large. However, the corresponding voids ratio was decreased from 0.361 to 0.192

by 46.8%, which is very large. Such decrease in voids ratio would substantially

reduce the amount of water needed to fill the voids and increase the amount of

excess water available for forming water films.

4.5.2 Water film thickness

The WFT results are listed in the last column of Table 4.1 and plotted

against the W/S ratio for different FAM and CSF contents in Figure 4.2. All the

WFT - W/S ratio curves are straight lines because the WFT is a linear function of

the W/S ratio. From the figure, it can be seen that some of mortar mixes have

negative WFT values. When the WFT is negative, it no longer has the physical

meaning as the average thickness of water films coating the solid particles. A

negative WFT value indicates that the amount of water in the mortar mix is not

- 55 -

sufficient to fill the voids between the solid particles, leading to the entrapment of

air in the mortar.

With only OPC and no FAM or CSF added, the WFT varied from -0.182

µm at W/S ratio = 0.26 to 0.429 µm at W/S ratio = 0.60. With 20% FAM added,

the WFT was increased to -0.066 µm at W/S ratio = 0.26 and to 0.452 µm at W/S

ratio = 0.60. With 40% FAM added, the WFT was further increased to 0.004 µm

at W/S ratio = 0.26 and to 0.454 µm at W/S ratio = 0.60. Hence, the addition of

FAM up to 40% would significantly increase the WFT within the range of W/S

covered in this study. The increase in WFT was generally larger at lower W/S

ratio. This was because of the proportionally larger increase in excess water at

lower water content. On the other hand, with 10% CSF but no FAM added, the

WFT was increased to 0.007 µm at W/S ratio = 0.26 but decreased to 0.325 µm at

W/S ratio = 0.60. The addition of CSF increased the WFT at W/S ratio ≤ 0.45 but

decreased the WFT at W/S ratio ≥ 0.50. Hence, the addition of CSF up to 10% did

not always increase the WFT. This was because the CSF has a very large specific

surface area and its addition has dramatically increased the solid surface area to be

coated with water films and thus thinned down the WFT, especially at high W/S

ratio when the proportional increase in excess water was relatively small. Lastly,

with CSF already added, the addition of FAM up to 40% increased the WFT at

W/S ratio < 0.40 but decreased the WFT at W/S ratio > 0.40. This was because of

the proportionally larger increase in excess water at lower W/S ratio and smaller

increase in excess water at higher W/S ratio.

From the above, it is obvious that the FAM, which is finer than OPC but

not as fine as CSF, is effective in increasing the WFT over a wider range of W/S

ratio than the CSF, while the CSF, which is the finest of all, is more effective in

increasing the WFT at very low W/S ratio. Hence, the optimum FAM and CSF

contents for maximum WFT are dependent on the W/S ratio. In this case, ternary

blending of OPC with 40% FAM and 10% CSF is the optimum at W/S ratio <

0.30 but binary blending of OPC with 40% FAM is the optimum at W/S ratio >

0.30.

- 56 -

4.5.3 Flow spread

The flow spread results are listed in the second column of Table 4.2 and

plotted against the W/S ratio in the upper half of Figure 4.3. It can be seen that the

mortar samples started to flow at different W/S ratios. For instance, the mortar

mix with 40% FAM and 10% CSF started to flow at a W/S ratio as low as 0.172

while the one with no FAM and CSF started to flow only when the W/S ratio was

increased to 0.387. Overall, blending of OPC with FAM and/or CSF had

significantly lowered the W/S ratio at which the mortar started to flow and

increased the flow spread at the same W/S ratio. Evidently, ternary blending with

both FAM and CSF is more effective than binary blending with FAM or CSF in

increasing the flow spread and lowering the W/S ratio needed for a given flow

spread requirement.

To study the role of the WFT, the flow spread results are plotted against

the WFT in the lower half of Figure 4.3. It is noted that all the mortar samples

started to flow at almost the same WFT of around 0 to 0.05 µm. This finding

agrees closely with the value of 0.025 µm obtained by Hunger and Brouwers

[2009] and Quercia et al. [2012]. Using this WFT value as the minimum WFT

needed for a mortar to flow, the water demand of mortar or concrete may be taken

as the volume of water needed to fill the voids and provide the minimum WFT.

Overall, as the WFT increased, the flow spread also increased at a gradually

decreasing rate. When the WFT increased to beyond 0.40 µm, the flow spread

further increased only marginally.

Among the mortar mixes with binary blended OPC + FAM, the flow

spread was larger at higher FAM content even at the same WFT, indicating that

the addition of FAM improves the flow spread not only by increasing the WFT

but also by some other beneficial effects. It is postulated herein that such

beneficial effects may be attributed to the ball bearing effect of the FAM particles

located within the gaps between the larger size particles (the cement grains and

fine aggregate particles). Being spherical in shape, the FAM particles can roll like

ball bearings to reduce the inter-particle friction between the larger size particles.

- 57 -

Comparing the flow spread - WFT curves of the various mortar mixes with

or without CSF, it can be seen that the flow spread of a mortar containing CSF

was substantially larger than that of a similar mortar containing no CSF even at

the same WFT. Particularly, at the same WFT, the flow spread of mortar added

with 10% CSF but no FAM was significantly larger than that of mortar added

with 40% FAM but no CSF. Hence, the addition of CSF appears to be more

effective than the addition of FAM in increasing the flow spread at the same WFT.

Since the CSF particles are also spherical, it may be postulated that one of the

contributing factors is the ball bearing effect of the CSF particles. Moreover, the

CSF particles are ultrafine and tend to move together with the water to form a

water-CSF slurry coating the larger size particles. Since the water-CSF slurry has

a larger volume than the water itself, the presence of CSF would increase the

thickness of the slurry so as to provide better lubrication for the water-solid

mixture.

To study the combined effects of the WFT and the SCM added, multi-

variable regression analysis has been carried out to derive the best-fit curves for

the flow spread - WFT relation. The best-fit curves so obtained are plotted

alongside the data points and the equation is also presented in the graph. A very

high R2 value of 0.964 has been achieved, indicating that the flow spread is highly

related to the WFT, FAM content and CSF content. Generally, the WFT has the

greatest effect and is therefore considered a key factor. At the same WFT, the

FAM and CSF have additional beneficial effects on the flow spread as observed in

the graph.

4.5.4 Flow rate

The flow rate results are listed in the third column of Table 4.2 and plotted

against the W/S ratio in the upper half of Figure 4.4. Overall, blending of OPC

with FAM and/or CSF had significantly lowered the W/S ratio at which the mortar

started to flow and increased the flow rate at the same W/S ratio. Moreover,

ternary blending with both FAM and CSF is more effective than binary blending

with FAM or CSF in increasing the flow rate and lowering the W/S ratio needed

for a given flow rate requirement.

- 58 -

The flow rate results are plotted against the WFT in the lower half of

Figure 4.4. The mortar did not flow when the WFT was negative or very small,

and started to flow at a WFT of around 0 to 0.10 µm, which agrees with the value

of 0.025 µm obtained by Hunger and Brouwer [2009] and Quercia et al. [2012].

Due probably to the ball bearing effects of the spherical SCM particles, a mortar

sample with a higher SCM content generally started to flow at a slightly smaller

WFT. Overall, the flow rate increased with the WFT at an increasing rate and at

the same WFT, the flow rate was higher at higher FAM and/or CSF contents.

Hence, the addition of FAM and/or CSF not only increased the WFT but also

increased the flow rate at the same WFT. Such phenomena may be explained by

the same reasons given in the previous section for the effects of adding FAM

and/or CSF on the flow spread.



the flow rate - WFT relation. The best-fit curves so obtained are plotted alongside

the data points and the equation is also presented in the graph. A very high R2

value of 0.991 has been achieved, indicating that the flow rate is highly related to

the WFT, FAM content and CSF content. Generally, the WFT has the greatest

effect and is therefore considered a key factor. At the same WFT, the FAM and

CSF have additional beneficial effects on the flow rate as observed in the graph.

4.5.5 Yield stress

The yield stress results are listed in the fourth column of Table 4.2 and

plotted against the W/S ratio and the WFT in Figure 4.5. At low W/S ratio, the

torque required to shear the mortar had sometimes exceeded the capacity of the

rheometer, causing the yield stress to be undetermined (shown as a hyphen in

Table 4.2). From the yield stress - W/S ratio curves, it is evident that blending of

OPC with FAM and/or CSF had significantly lowered the yield stress, especially

at low W/S ratio. From the yield stress - WFT curves, it is found that even at the

same WFT, the yield stress was lower at higher FAM or CSF contents. Apparently,

ternary blending with both FAM and CSF is more effective than binary blending

with either FAM or CSF in reducing the yield stress at the same WFT. These

- 59 -

observed phenomena may be explained by the same reasons given in the previous

section on flow spread.



the yield stress - WFT relation. The best-fit curves so obtained are plotted


high R2 value of 0.983 has been achieved, indicating that the yield stress is highly

related to the WFT, FAM content and CSF content. Generally, the WFT has the

greatest effect and is therefore considered a key factor. At the same WFT, the

FAM and CSF have the additional effects of decreasing the yield stress.

4.5.6 Apparent viscosity

The apparent viscosity results are listed in the fifth column of Table 4.2

and plotted against the W/S ratio and WFT in Figure 4.6. Some apparent viscosity

results (each shown as a hyphen) were not obtained because the torque required to

shear the mortar had exceeded the capacity of the rheometer. The apparent

viscosity - W/S ratio curves show that blending of OPC with FAM and/or CSF

had significantly lowered the apparent viscosity, especially at low W/S ratio. On

the other hand, the apparent viscosity - WFT curves reveal that even at the same

WFT, a mortar with high FAM or CSF contents had lower apparent viscosity.

Overall, ternary blending with both FAM and CSF appears to be more effective

than binary blending with either FAM or CSF in reducing the apparent viscosity at

the same WFT. These phenomena may be explained by the same reasons given in

the previous section on flow spread.



the apparent viscosity - WFT relation. The best-fit curves so obtained are plotted


high R2 value of 0.948 has been achieved, indicating that the apparent viscosity is

highly related to the WFT, FAM content and CSF content. Generally, the WFT

has the greatest effect and is therefore considered the governing factor. At the

- 60 -

same WFT, the FAM and CSF have the additional effects of decreasing the

apparent viscosity.

4.5.7 Adhesiveness

The results of the stone rod adhesion test are listed in the sixth column of

Table 4.2 and plotted against the W/S ratio in the upper half of Figure 4.7. From

the figure, it can be seen that all the adhesiveness - W/S ratio curves follow the

trend that the adhesiveness first increased from a very low level as the W/S ratio

increased and then, after the adhesiveness had reached a certain peak value at an

optimum W/S ratio, the adhesiveness decreased as the W/S ratio further increased.

So, the water added to the mortar may be beneficial or detrimental to the

adhesiveness and to achieve maximum adhesiveness, the water content should not

be too low or too high. Generally, the optimum W/S ratio was lower and the

maximum adhesiveness was higher at higher FAM or CSF contents.

To investigate the role of the WFT, the adhesiveness results are plotted

against the WFT in the lower half of Figure 4.7. Compared to the adhesiveness -

W/S ratio curves, the adhesiveness - WFT curves are closer to each other. This

demonstrates that the WFT plays a more important role in the adhesiveness of

mortar than the W/S ratio. Generally, the maximum adhesiveness occurred at an

optimum WFT of around 0 to 0.10 µm. In fact, when the WFT was negative, the

mortar mix appeared to be too dry to have any adhesion. On the other hand, when

the WFT was larger than 0.10 µm, the mortar mix appeared to be too wet to

adhere to the stone rods. It may thus be concluded that the adhesiveness is highest

when the mortar mix has the right wetness, as indicated by the WFT falling within

0 to 0.10 µm. Although there has been little research on this topic, increasing the

adhesiveness of mortar should be an effective method to produce high-build

mortar for rendering and concrete repair, and the mortar portion of concrete

requiring high cohesiveness (which is highly dependent on the ability of the

mortar to adhere to the aggregate particles).

- 61 -

4.5.8 Cube strength

The 28-day cube strength results are listed in the last column of Table 2

and plotted against the W/CM ratio by mass in the upper half of Figure 4.8. Each

cube strength result presented therein is the average of the three cubes casted from

the same batch and tested at the same time. These results show that the cube

strength increased as the W/CM ratio decreased until the W/CM ratio reached an

optimum value, and then the cube strength turned to decrease as the W/CM ratio

further decreased. The optimum W/CM ratio for maximum strength was lower

and the maximum strength was higher at higher SCM contents. With no FAM or

CSF added, the maximum cube strength attained was 96.2 MPa at a W/CM ratio

of 0.289. With 40% FAM and 10% CSF added, the maximum cube strength was

increased to 129.6 MPa at a lower W/CM ratio of 0.180. Actually, at the same and

relatively high W/CM ratio, the addition of FAM had little effect on the strength

and the addition of CSF only slightly increased the strength. It was the reduction

in optimum W/CM ratio that led to the higher maximum strength at higher FAM

or CSF contents.

The 28-day cube strength results are plotted against the WFT in the lower

half of Figure 4.8. Here, a clear trend is revealed that the maximum strength

always occurred at a WFT of around 0.00 to 0.05 µm. When the WFT was

negative, the water was not sufficient to fill the voids and in such case, air was

entrapped inside the voids no matter how hard the mortar was compacted,

resulting in a relatively low strength. When the WFT was positive and higher than

0.05 µm, the strength decreased as the WFT increased because of the

corresponding increase in W/CM ratio. Hence, there is a fixed optimum WFT for

maximum strength regardless of the FAM and CSF contents. The optimum WFT

is positive but very close to zero, in which case, the water is slightly more than

sufficient to fill the voids.

- 62 -

4.6 Concurrent Strength-Flowability Performance

The concurrent strength - flowability performance of the mortar mixes is

illustrated by plotting the 28-day cube strength against the flow spread and flow

rate in Figure 4.9. From the cube strength - flow spread curves, it is evident that

the addition of 20% FAM and/or 10% CSF would shift the performance curve

upwards and to the right. This indicates that the addition of 20% FAM and/or 10%

CSF would increase the strength at the same flow spread, increase the flow spread

at the same strength, or increase both the strength and the flow spread. However,

the addition of 40% FAM makes little difference to the concurrent performance

compared with the addition of 20% FAM. In other words, the addition of more

than 20% FAM offers little benefit to the concurrent cube strength - flow spread

performance. Likewise, from the cube strength - flow rate curves, it can be seen

that the addition of 20% FAM and/or 10% CSF would improve the concurrent

cube strength - flow rate performance but the addition of more than 20% FAM

offers little benefit to the concurrent cube strength – flow rate performance.

4.7 Conclusions

From the previous discussions, the following conclusions can be drawn:

(1) Blending of OPC with fine SCM can effectively improve the packing

density of the solid particles in mortar. Relatively, ternary blending with

both FAM and CSF is more effective than binary blending with either

FAM or CSF in increasing the packing density. Moreover, binary blending

with CSF is more effective than binary blending with FAM because CSF

has higher fineness.

(2) The effect of adding fine SCM on the WFT varies with the W/S ratio. At

low W/S ratio, the proportional increase in excess water due to the increase

in packing density is larger and thus the increase in WFT would be larger.

At high W/S ratio, the WFT would increase less and might even decrease.

(3) The optimum blending for maximum WFT is dependent on the W/S ratio.

- 63 -

At W/S ratio < 0.30, ternary blending with 40% FAM and 10% CSF is

optimum. At W/S ratio > 0.30, binary blending with 40% FAM is

optimum.

(4) Correlations of the measured flow spread, flow rate, yield stress and

apparent viscosity to the WFT yielded very high R2 values of well above

0.9, indicating that the WFT plays an important role in the rheology of

mortar.

(5) The addition of FAM and/or CSF has some additional effects on the

rheology of mortar, which are not reflected by the WFT. This is

considered to be caused by the ball bearing effects of the FAM and CSF

particles and the formation of water-CSF slurry coating the larger size

particles.

(6) The addition of FAM and/or CSF has complicated effects on the

adhesiveness. Nevertheless, it is evident that the adhesiveness is highest

when the mortar is neither too dry nor too wet, as indicated by its WFT,

taken as a measure of wetness, falling within 0 to 0.10 µm.

(7) The optimum W/CM ratio for maximum strength is generally lower at

higher FAM and CSF contents. Hence, the addition of FAM and CSF

would allow the W/CM ratio to be lowered to increase the strength.

Although the optimum W/CM ratio for maximum strength varies with the

FAM and CSF contents, the optimum WFT for maximum strength is

constant at around 0 to 0.05 µm.

(8) From plots of concurrently achieved strength and flowability, it can be

seen that the addition of FAM and/or CSF would increase the strength at

the same flowability, increase the flowability at the same strength or

increase both the strength and the flowability. Considering both the

strength and flowability, ternary blending with 20% FAM and 10% CSF

seems to be the optimum.

- 64 -

Tables

Table 4.1 Mix proportions, packing density and WFT


mass


Packing density (Voids ratio)

WFT (µm)

OPC FAM CSF FA* Water SP

M-0-0-0.6 0.193 1061 0 0 1146 202 12

0.735 (0.361)

-0.182

M-0-0-0.7 0.225 1026 0 0 1108 229 11 -0.107

M-0-0-0.8 0.257 993 0 0 1073 253 11 -0.029

M-0-0-0.9 0.289 962 0 0 1040 276 11 0.047

M-0-0-1.0 0.321 934 0 0 1009 298 10 0.124

M-0-0-1.2 0.386 881 0 0 952 338 10 0.276

M-0-0-1.4 0.450 834 0 0 901 373 9 0.429

M-20-0-0.6 0.200 849 172 0 1146 198 17

0.768 (0.302)

-0.066

M-20-0-0.7 0.234 821 166 0 1108 225 16 -0.001

M-20-0-0.8 0.267 795 161 0 1073 250 16 0.063

M-20-0-0.9 0.301 770 156 0 1040 273 15 0.128

M-20-0-1.0 0.334 747 151 0 1009 295 15 0.193

M-20-0-1.2 0.401 705 143 0 952 334 14 0.322

M-20-0-1.4 0.468 667 135 0 901 370 13 0.452

M-40-0-0.5 0.174 659 356 0 1187 166 22

0.797 (0.255)

-0.052

M-40-0-0.6 0.209 637 344 0 1146 194 22 0.004

M-40-0-0.7 0.244 616 332 0 1108 221 21 0.061

M-40-0-0.8 0.278 596 322 0 1073 246 20 0.117

M-40-0-0.9 0.313 577 312 0 1040 269 20 0.173

M-40-0-1.0 0.348 560 302 0 1009 291 19 0.229

M-40-0-1.2 0.417 528 285 0 952 331 18 0.342

M-40-0-1.4 0.487 500 270 0 901 367 17 0.454

* FA means fine aggregate.

- 65 -

Table 4.1 Mix proportions, packing density and WFT (Continued)


mass



WFT (µm)

OPC FAM CSF FA* Water SP

M-0-10-0.5 0.166 989 0 78 1187 165 23

0.800 (0.250)

-0.032

M-0-10-0.6 0.199 955 0 75 1146 194 22 0.007

M-0-10-0.8 0.265 894 0 70 1073 245 21 0.087

M-0-10-1.0 0.331 840 0 66 1009 291 20 0.166

M-0-10-1.2 0.397 793 0 62 952 331 18 0.246

M-0-10-1.4 0.464 750 0 59 901 367 17 0.325

M-20-10-0.4 0.138 797 184 80 1230 130 29

0.821 (0.218)

-0.035

M-20-10-0.5 0.172 769 178 78 1187 161 28 -0.002

M-20-10-0.6 0.207 743 172 75 1146 190 27 0.030

M-20-10-0.8 0.276 695 161 70 1073 241 25 0.095

M-20-10-1.0 0.345 654 151 66 1009 287 24 0.160

M-20-10-1.2 0.414 617 143 62 952 327 23 0.225

M-20-10-1.4 0.482 584 135 59 901 363 21 0.290

M-40-10-0.4 0.144 569 369 80 1230 126 34

0.839 (0.192)

-0.013

M-40-10-0.5 0.180 549 356 78 1187 157 33 0.015

M-40-10-0.6 0.216 530 344 75 1146 186 32 0.042

M-40-10-0.8 0.287 497 322 70 1073 238 30 0.097

M-40-10-1.0 0.359 467 302 66 1009 284 28 0.152

M-40-10-1.2 0.431 440 285 62 952 324 27 0.207

M-40-10-1.4 0.503 417 270 59 901 360 25 0.261


- 66 -


Mix no. Flow

spread (mm)

Flow rate

(ml/s)

Yield stress (Pa)

Apparent viscosity

(Pas)

Adhesive- ness (g)

28-day cube

strength (MPa)

M-0-0-0.6 0.0 0.0 - - 0.0 8.7

M-0-0-0.7 0.0 0.0 - - 0.1 57.5

M-0-0-0.8 0.0 0.0 - - 0.2 95.0

M-0-0-0.9 5.5 0.0 - - 2.0 96.2

M-0-0-1.0 92.0 23.0 47.2 19.5 39.0 89.1

M-0-0-1.2 182.0 192.2 20.0 6.9 15.7 71.5

M-0-0-1.4 210.0 405.0 6.0 3.1 7.1 61.0

M-20-0-0.6 0.0 0.0 - - 0.1 48.2

M-20-0-0.7 0.0 0.0 - - 0.5 116.2

M-20-0-0.8 57.5 0.0 - - 23.4 109.6

M-20-0-0.9 111.5 67.5 32.0 14.0 48.2 107.3

M-20-0-1.0 151.0 151.2 19.9 8.8 39.5 91.3

M-20-0-1.2 210.0 378.0 7.0 3.6 19.0 81.9

M-20-0-1.4 225.0 691.5 3.5 1.7 14.1 66.6

M-40-0-0.5 0.0 0.0 - - 0.1 117.3

M-40-0-0.6 0.0 0.0 - - 0.8 121.9

M-40-0-0.7 62.5 59.7 41.0 22.3 57.5 109.0

M-40-0-0.8 121.5 143.5 20.4 11.1 43.3 100.9

M-40-0-0.9 153.5 241.3 14.1 8.1 28.0 93.0

M-40-0-1.0 191.0 354.4 9.6 5.0 19.9 84.3

M-40-0-1.2 238.5 590.6 3.9 1.9 12.9 67.7

M-40-0-1.4 247.5 859.1 3.2 1.2 10.3 55.4

- 67 -


(Continued)

Mix no. Flow

spread (mm)

Flow rate

(ml/s)

Yield stress (Pa)

Apparent viscosity

(Pas)

Adhesive- ness (g)

28-day cube

strength (MPa)

M-0-10-0.5 0.0 - \ \ 0.0 98.6

M-0-10-0.6 0.0 - \ \ 0.1 117.6

M-0-10-0.8 89.0 49.6 23.7 21.3 39.6 105.9

M-0-10-1.0 172.0 157.5 9.2 6.6 22.0 102.2

M-0-10-1.2 243.0 289.3 1.7 3.0 15.0 87.8

M-0-10-1.4 269.0 534.9 1.7 1.2 10.8 73.6

M-20-10-0.4 0.0 - \ \ 0.0 101.1

M-20-10-0.5 10.0 - \ \ 2.1 118.5

M-20-10-0.6 61.0 17.2 38.7 23.4 56.9 128.1

M-20-10-0.8 157.5 117.6 9.1 7.3 24.2 105.3

M-20-10-1.0 215.0 234.3 4.3 4.1 17.2 85.3

M-20-10-1.2 250.0 363.5 3.2 2.8 8.7 83.5

M-20-10-1.4 275.5 578.6 1.6 0.7 8.6 67.5

M-40-10-0.4 7.0 - \ \ 0.1 119.8

M-40-10-0.5 57.5 - 53.4 32.1 59.7 129.6

M-40-10-0.6 95.0 24.3 27.3 26.7 51.8 117.4

M-40-10-0.8 196.5 140.3 4.9 6.6 18.6 95.9

M-40-10-1.0 237.0 265.0 2.7 4.0 10.0 88.1

M-40-10-1.2 260.0 457.3 1.9 1.4 5.7 70.2

M-40-10-1.4 281.5 603.2 1.6 0.7 5.2 53.4

- 68 -

Figures

0

20

40

60

80

100

0.1 1 10 100 1000 10000

Particle size (µm)

Per

cent

age

pass

ing

(%

) .

Figure 4.1 Particle size distributions of CSF, FAM, OPC and fine aggregate

FAM OPC CSF fine aggregate

- 69 -

-0.20

-0.10

0.00

0.10

0.20

0.30

0.40

0.50

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

WF

T (

µm)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

Figure 4.2 Variation of WFT with W/S ratio at different FAM and CSF contents

- 70 -

y = A + BeCx1

A = 402-0.627x2-0.234x3

B = -399+0.225x2-1.149x3 C = -5.412-0.212x2-0.009x3 R

2 = 0.932

0

50

100

150

200

250

300

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

Flo

w s

prea

d (m

m)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

50

100

150

200

250

300

350

-0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Flo

w s

prea

d (m

m)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

Note: x1 is the WFT (µm); x2 is the FAM content (%); x3 is the CSF content (%)

Figure 4.3 Flow spread versus W/S ratio and WFT

y = A + BeCx1 A = 252+0.512x2+5.789x3

B = -260-0.328x2-2.661x3 C = -3.624-0.045x2-0.157x3 R

2 = 0.964

- 71 -

0

200

400

600

800

1000

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

Flo

w r

ate

(ml/

s)FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

200

400

600

800

1000

-0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Flo

w r

ate

(ml/

s)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%


Figure 4.4 Flow rate versus W/S ratio and WFT

y = A + Bx1+C x12

A = -50

B = 335+23x2+76x3 C = 1774 R

2 = 0.991

- 72 -

0

10

20

30

40

50

60

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

Yie

ld s

tres

s (P

a)FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

10

20

30

40

50

60

70

-0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Yie

ld s

tres

s (P

a)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%


Figure 4.5 Yield stress versus W/S ratio and WFT

y = A + BeCx1 A = 3.730-0.020x2-0.200x3

B = 96-0.238x2-1.495x3 C = -6.396-0.160x2-1.086x3

R2 = 0.983

- 73 -

0

10

20

30

40

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

App

aren

t vi

scos

ity

(Pas

)FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

10

20

30

40

-0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

App

aren

t vi

scos

ity

(Pas

)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%


Figure 4.6 Apparent viscosity versus W/S ratio and WFT

y = A + BeCx1 A = 0.177

B = 41 C = -5.956-0.112x2-0.492x3

R2 = 0.948

- 74 -

0

20

40

60

0.15 0.25 0.35 0.45 0.55 0.65

W/S ratio by volume

Adh

esiv

enes

s (g

)FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

20

40

60

-0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

Adh

esiv

enes

s (g

)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

Figure 4.7 Adhesiveness versus W/S ratio and WFT

- 75 -

0

20

40

60

80

100

120

140

0.10 0.20 0.30 0.40 0.50 0.60

W/CM ratio by mass

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

20

40

60

80

100

120

140

-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50

WFT (µm)

28-d

ay c

ube

stre

ngth

(M

Pa)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

Figure 4.8 Cube strength versus W/S ratio and WFT

- 76 -

0

20

40

60

80

100

120

140

0 50 100 150 200 250 300

Flow spread (mm)

28-d

ay C

ube

stre

ngth

(M

pa)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

0

20

40

60

80

100

120

140

0 200 400 600 800 1000

Flow rate (ml/s)

28-d

ay C

ube

stre

ngth

(M

pa)

FAM= 0%, CSF= 0%

FAM=20%, CSF= 0%

FAM=40%, CSF= 0%

FAM= 0%, CSF=10%

FAM=20%, CSF=10%

FAM=40%, CSF=10%

Figure 4.9 Concurrent cube strength and flowability performance

- 77 -

CHAPTER 5

TERNARY CEMENTITIOUS SYSTEM CONTAINING

SUPERFINE CEMENT AND CONDENSED SILICA FUME

5.1 Introduction

The strength and durability of concrete can be greatly improved by

lowering the water/cementitious materials (W/CM) ratio [Aїtcin, 1998]. However,

since the water added must be more than sufficient to fill the voids in the bulk

volume of the cementitious materials, there is a limit to which the W/CM ratio can

be lowered. Likewise, the dimensional stability of concrete can be enhanced and

the cement consumption and carbon footprint can be reduced by decreasing the

cement paste volume [Kwan, 2003]. But, as the cement paste volume must be

more than sufficient to fill the voids in the bulk volume of the aggregate, there is

also a limit to which the cement paste volume can be decreased. Hence,

maximization of particle packing density so as to reduce the volume of voids to be

filled is the key to the production of high-performance concrete (HPC). For this

purpose, the addition of supplementary cementitious materials (SCM) finer than

cement to fill into the voids between cement grains is particularly effective. In fact,

concrete produced with SCM added is often found to perform better in terms of

workability, strength and durability [Khatri et al., 1995, Aїtcin, 1998, Rizwan and

Bier, 2009].

Although the use of SCM is already quite common, the effects of SCM are

still not fully understood, as reflected by the widely different and even

contradictory results reported by different researchers. Take condensed silica

fume (CSF), which is considered one of the most effective SCM [F.I.P.

Commission on Concrete, 1988], as an example. Due to its ultra-high fineness and

high pozzolanic activity, the addition of CSF can significantly increase the

concrete strength. But, the effects of CSF on the fresh properties of cement

- 78 -

paste/mortar/concrete are fairly complicated and up to now no consensus has been

reached. Kohno and Komatsu [1986] reported that at the same W/CM ratio, a

mortar containing CSF would have a smaller flow compared with a plain mortar

containing no CSF. However, Duval and Kadri [1998] demonstrated that cement

replacement by CSF up to 10% by mass has no adverse effect on concrete

workability. Similarly, Zhang and Han [2000] showed that the yield stress and

viscosity of a cement paste could be decreased by adding CSF to replace 10% by

mass of cement. More recently, Artelt and Garcia [2008] examined several mortar

mixes with and without CSF and concluded that the presence of CSF would

impair the flowability as evidenced by the smaller flow spread and longer flow

time obtained for the CSF mortars.

The author is of the view that the addition of fine SCM has two major

effects and it is the combined action of these two effects that governs the rheology

of cement paste/mortar/concrete. First, the fine SCM particles would fill into the

voids between cement grains to increase the packing density of the particle system.

As a result, the volume of voids would decrease and the amount of water in excess

of that needed to fill the voids (i.e. the amount of excess water) for forming water

films coating the solid particles would increase. Such effect is beneficial to the

flowability of the water-solid mixture. Another effect is that with fine SCM added,

the total surface area of the solid particles would increase so that with the same

amount of excess water available, the average thickness of the water films coating

the solid particles would decrease. This is in agreement with Claisse et al. [2001],

who pointed out that the specific surface area of the solid particles has great

influence on the rheology of cement paste, and Ferraris et al. [2001], who

suggested that the increase in packing density and the increase in solid surface

area have opposite effects on the rheology of cement paste. Consequently, the net

influence of adding fine SCM is dependent on the relative magnitudes of the two

opposite effects.

Dated back to the 1960s’, Powers [1968] proposed that it is the excess

paste (paste in excess of that needed to fill the voids between aggregate particles)

that gives the mortar or concrete workability. Similarly, Helmuth [1980]

suggested that it is the excess water (water in excess of that needed to fill the

- 79 -

voids between cement grains) that forms water films coating the cement grains

and it should be the thickness of the water films that governs the consistence of

cement paste. Zhang et al. [1996] found that the addition of fine pozzolanic

material could increase the packing density and thereby decrease the amount of

filling water in the voids and increase the amount of excess water in the surface

layers. Later, Kwan and his research team [Kwan and Wong, 2008a, Wong and

Kwan, 2008a] proposed that the average water film thickness (WFT) may be

evaluated indirectly as the excess water to solid surface area ratio. Hence, by

measuring the packing density and solid surface area of the solid particles, the

WFT can be determined without direct measurement.

However, the packing density of fine solid particles is not easy to measure,

because of agglomeration caused by the presence of inter-particle forces [Yu et al.,

1997]. The existing dry packing methods of measuring packing density under dry

condition have the major problems that the measured results are sensitive to the

amount of compaction applied [Svarovsky, 1987] and that they cannot take into

account the effects of water and superplasticizer, which have significant effects on

the packing density. Without accurate measurement of packing density, it is

impossible to precisely determine the WFT for investigating the effects of WFT.

In other words, the lack of a proper method for measuring the packing density of

fine solid particles has been the main hurdle in the study of WFT.

Recently, the author’ research team has developed a wet packing method

for direct measurement of the packing densities of cementitious materials [Wong

and Kwan, 2008b], fine aggregate [Fung et al., 2009] and cementitious materials

plus fine aggregate [Kwan and Fung, 2009], and by comparing with theoretical

results based on existing packing models, verified the accuracy and applicability

of this wet packing method [Kwan and Fung, 2009, Wong and Kwan, 2008c].

With the packing density directly measured and then the WFT determined as the

excess water to solid surface area ratio, the author’ research team has

demonstrated that the WFT is the single most important parameter governing the

rheology of cement paste [Kwan and Wong, 2008a, Wong and Kwan, 2008a].

- 80 -

Herein, a comprehensive testing program was carried out to study the roles

of WFT in mortar mixes containing triple-blended cementitious materials. To

maximize the packing density of the cementitious materials, superfine cement

(SFC), which was finer than cement, was added to fill the voids between the

cement grains, and condensed silica fume (CSF), which had the highest fineness,

was added to fill the voids between the other cementitious materials. A number of

mortar samples with different SFC and CSF contents and different W/CM ratios

were made for packing density, rheology, adhesiveness and strength

measurements. From the test results, the rheological properties, adhesiveness and

strength of the mortar samples were correlated to the WFT for in-depth analysis of

the roles of WFT in the fresh and hardened properties of mortar.

5.2 Materials

The cement used was an ordinary Portland cement (OPC) obtained from

the local market. It was of strength class 52.5N and had been tested to comply

with BS EN 197-1:2000. The SFC and CSF used were imported from Europe.

According to the supplier, the SFC was a slag-cement containing 80% slag and

20% cement. On the other hand, the CSF had been tested by the supplier to

comply with ASTM C 1240-03. The fine aggregate used was a local crushed

granite rock fine with a maximum size of 1.18 mm and a water absorption of 1.6%

by mass. The relative densities of the OPC, SFC, CSF and fine aggregate had been

measured in accordance with BS EN 196-6 2010 as 3.11, 2.94, 2.20 and 2.48

respectively. A laser diffraction particle size analyzer was used to measure the

particle size distributions of the materials and the results are plotted in Figure 5.1.

Based on their particle size distributions, the specific surface areas of the OPC,

SFC, CSF and fine aggregate were calculated as 1.12×106 m2/m3, 2.29×106 m2/m3,

1.33×107 m2/m3 and 2.1×105 m2/m3, respectively. The SP employed was a

polycarboxylate-based type with a solid mass content of 20% and a relative

density of 1.03.

- 81 -


To study the roles of WFT in fresh and hardened properties of mortar, an

experimental program was carried out. Three different SFC contents, namely 0%,

10% and 20%, and two different CSF contents, namely 0% and 10%, each

expressed as a percentage by volume of the total cementitious materials, were

adopted for the design of the mortar samples. To exclude the effect of variation in

fine aggregate content, the total cementitious materials to fine aggregate ratio was

fixed at 0.75 by volume. The W/CM ratio by volume was varied from 0.5 to 1.2.

In total, 27 mortar samples were produced for testing. The mix proportions of the

mortar samples are summarized in Table 5.1. Each mortar sample is assigned a

designation of M-X-Y-Z, where M denotes mortar, X and Y denote the SFC and

CSF contents, respectively, and Z denotes the W/CM ratio by volume. For

reference, the W/CM ratio by mass of each mortar sample is listed in the second

column of Table 5.1.

A superplasticizer (SP) was added to each mortar sample. Since SP is a

surface reactant and it is the SP dosage per solid surface area that actually governs

the effectiveness of the SP [Wong and Kwan, 2008a, Kwan et al., 2012], the SP

dosage was determined according to the total surface area of the solid particles in

the mortar. Before setting the SP dosage to be used, trial cement paste mixing

using different SP dosages was carried out and it was found that the saturation

dosage (the dosage beyond which further addition of the SP yields little further

increase in flowability) was 2.6×10-7 kg/m2 of the solid surface area. Hence, the

SP dosage in terms of liquid mass of SP per solid surface area was set as 2.6×10-7

kg/m2 for all the mortar samples. It is noteworthy that since SFC and CSF have

higher fineness, the SP dosage per mass of cementitious materials was higher at

higher SFC and/or CSF contents.

The experimental program consisted of three parts. The first part was to

measure the packing density of the solid particles in each mortar sample by a wet

packing method developed by the author’ research team. In the second and third

parts, the fresh and hardened properties of mortar samples were measured. Each

- 82 -

mortar sample was produced using a standard mixer by first adding all the water

and SP to the mixer and then adding the solid materials bit by bit to the mixer

while mixing. This mixing procedure has been found to be more effective than the

conventional mixing procedure of adding all the water and solid materials to the

mixer in one single batch, especially when the water content is low and/or

ultrafine filler is added [Wong and Kwan, 2008b]. All of the mixing and testing

procedures were carried out in a laboratory maintained at a temperature of 24 ± 2

ºC.

5.4 Test Methods

Measurement of Packing Density

The six mixes of solid samples in the mortar samples were subjected to the

packing density test. The details of the test procedures can be referred to section

2.1 in Chapter 2.

Determination of Water Film Thickness








Chapter 2.


- 83 -




Chapter 2.




2.3, 2.4 2.5 and 2.6 in Chapter 2.






5.5.1 Packing density and water film thickness

The measured packing densities of the mortar mixes are tabulated in the

ninth column of Table 5.1. It can be seen from these results that without the

addition of any SFC or CSF, the solid mix of OPC and fine aggregate was

measured to have a packing density of 0.737. With SFC but no CSF added, the

packing density was increased to 0.749 at 10% SFC content and to 0.755 at 20%

SFC content. With CSF but no SFC added, the packing density was increased to

0.780 at 10% CSF content. This demonstrates that the SFC and CSF, when added

individually, could effectively increase the packing density of the solid particles in

mortar. Compared to the SFC, the CSF was more effective due to its higher

fineness.

- 84 -

The results also show that with both SFC and CSF added, the packing

density of the solid particles could be further increased. With 10% CSF added, the

packing density was increased by 5.8% from 0.737 to 0.780, by 5.3% from 0.749

to 0.789, and by 6.0% from 0.755 to 0.800 at 0, 10% and 20% SFC contents,

respectively. Hence, regardless of the SFC content, the addition of 10% CSF

would increase the packing density by about 5.3% to 6.0%. On the other hand,

with 20% SFC added, the packing density was increased by 2.4% from 0.737 to

0.755 and by 2.6% from 0.780 to 0.800 at 0 and 10% CSF contents, respectively.

Hence, regardless of the CSF content, the addition of 20% SFC would increase the

packing density by about 2.4% to 2.6%. Overall, it is evident that blending OPC

with both SFC and CSF so that the voids between cement grains are successive

filled by finer and even finer particles is a more effective way of increasing the

packing density than just blending OPC with SFC or just blending OPC with CSF.

The voids ratios of the solid particles, as calculated by Equation (2.2), are

also tabulated in the ninth column of Table 5.1. It can be seen that single addition

of 20% SFC decreased the voids ratio from 0.357 to 0.325, single addition of 10%

CSF decreased the voids ratio from 0.357 to 0.282, whereas joint addition of 20%

SFC and 10% CSF decreased the voids ratio from 0.357 to 0.250. Compared to

the maximum increase in packing density of 8.5% which appears small, the

corresponding decrease in voids ratio of 30.0% is quite substantial. Such reduction

in voids ratio due to the addition of fine SCM would decrease the amount of water

needed to fill the voids between the solid particles so as to increase the amount of

excess water available for forming water films coating the solid particles in the

mortar.

The WFT of the mortar mixes are listed in the last column of Table 5.1 and

plotted against the W/S ratios for different SFC and CSF contents in Figure 5.2.

From the figure, it can be seen that at W/S ratio ≤ 0.45 (corresponding W/CM

ratio by mass ≤ 0.32), the addition of SFC and/or CSF would significantly

increase the WFT. However, at higher W/S ratio, the addition of SFC would

slightly increase the WFT but the addition of CSF would significantly decrease

the WFT. This phenomenon can be explained by Equation (2.5), which stipulates

that the WFT is determined by both the excess water ratio and the total solid

- 85 -

surface area. Generally, the proportional increase in excess water ratio due to the

addition of a SCM finer than cement is larger at lower W/S ratio and smaller at

higher W/S ratio whereas the proportional increase in total solid surface area is

larger when the SCM has higher fineness and smaller when the SCM has lower

fineness. If the proportional increase in excess water ratio is larger than the

proportional increase in total solid surface area, the WFT would increase,

otherwise, the WFT would decrease.

From the above results, it is evident that the SFC, which is finer than

cement but not as fine as CSF, is effective in increasing the WFT over a wider

range of water content but is not as effective as CSF at low water content. On the

other hand, the much finer CSF, which produces larger increases in packing

density and total solid surface area when added, is more effective than SFC in

increasing the WFT but only at low water content. Hence, the suitable amounts of

SFC and CSF to be added to increase the WFT are dependent on the water content.

5.5.2 Flow spread and flow rate

The flow spread and flow rate results are tabulated in the second and third

columns of Table 5.2 and plotted against the W/S ratio in Figure 5.3. As expected,

both the flow spread and flow rate generally increased with the W/S ratio.

However, the effects of SFC and CSF appeared to be fairly complicated. At W/S

ratio ≤ 0.45, the addition of SFC and/or CSF significantly increased the flow

spread and flow rate, but at W/S ratio > 0.45, the addition of SFC and/or CSF did

not always increase the flow spread and flow rate. Relatively, the CSF has greater

effects on the flowability than the SFC because of its larger effects on the WFT.

Overall, as depicted by the widely spaced flowability-W/S ratio curves for mortar

mixes containing different blends of SCM, it may be said that the flowability of

mortar is dependent not only on the W/S ratio, but also on the solid ingredient

contents in the mortar.

- 86 -

5.5.3 Yield stress and apparent viscosity

The yield stress and apparent viscosity results are tabulated in the fourth

and fifth columns of Table 5.2 and plotted against the W/S ratio in Figure 5.4. At

low W/S ratio, the torque required to shear the mortar sample had occasionally

exceeded the capacity of the rheometer causing the yield stress and apparent

viscosity of some mortar mixes to be undetermined (each shown in Table 5.2 by a

hyphen). From the curves plotted, it can be seen that both these two rheological

properties decreased as the W/S ratio increased and that the effects of SFC and

CSF were generally more significant at lower W/S ratio. On the whole, both

rheological properties varied more notably with the CSF content than with the

SFC content, as illustrated by the widely spaced curves plotted for different CSF

contents, which are obviously divided into two series, one series with no CSF

added and the other series with 10% CSF added. Hence, the rheological properties

of mortar are governed not only by the water content but also by the solid

ingredient contents in the mortar.

5.5.4 Adhesiveness

The stone rod adhesion test results are tabulated in the sixth column of

Table 5.2 and plotted against the W/S ratio in Figure 5.5. It is evident that

generally the adhesiveness varied with the W/S ratio in such a way that when the

W/S ratio was low, the adhesiveness increased as the W/S ratio increased, but

after reaching a certain peak value, the adhesiveness decreased as the W/S ratio

further increased. This implies that the water added to a mortar may have positive

or negative effect on the adhesiveness and there exists an optimum water content

depending on the mix proportions of the solid ingredients for maximum

adhesiveness.

5.5.5 Cube strength

The cube strength results are tabulated in the last column of Table 5.2.

Each cube strength result presented is the average of the three cubes cast and

tested at the same time. Generally, the cube strength varied with the W/CM ratio

- 87 -

in such a way that as the W/CM ratio decreased starting from a relatively high

value, the strength increased until a certain peak value was reached and then as the

W/CM ratio further decreased, the strength started to decrease. This was because

while the W/CM ratio was still high, the water was more than sufficient to fill the

voids and in such case, as usual, the strength increased as the W/CM ratio

decreased. However, when the W/CM ratio decreased to such level that the water

was no longer sufficient to fill the voids, air was entrapped inside the voids

causing the strength to decrease.

In this case, with no SCM added, the maximum cube strength achieved

was 91.4 MPa. With 10% SFC added, the maximum cube strength was increased

to 95.9 MPa. With 10% CSF added, the maximum cube strength was increased to

117.6 MPa. With 10% SFC and 10% CSF added together, the maximum cube

strength achieved was 116.6 MPa. These results show that the addition of CSF has

great effect on the strength whereas the addition of SFC has little effect especially

after CSF has been added. It is noteworthy that the optimum W/CM ratio for

maximum strength was lower at higher SFC and CSF contents.

5.6 Roles of Water Film Thickness

From the above, it is evident that the addition of SCM finer than cement

has significant effects on the packing density, solid surface area and WFT, which

in turn govern the fresh and hardened properties of mortar. The effects of SCM on

the packing density and solid surface area are not dependent on the W/S ratio but

the effects of SCM on the WFT are dependent on the W/S ratio. For this reason,

the net effects of SCM on the properties of mortar could vary with the water

content. Herein, it is suggested that the combined effects of the SCM content and

W/S ratio should be evaluated in terms of the WFT.

The WFT of the mortar samples, each determined as the respective excess

water to solid surface area ratio, are tabulated in the last column of Table 5.1. For

the mortar samples tested, the WFT ranged from -0.163 µm to 0.282 µm. It should

- 88 -

be noted that a negative WFT value indicates that the amount of water in the

mortar was not sufficient to fill the voids between the solid particles, leading to

the entrapment of air in the mortar. To study the roles of the WFT, the fresh and

hardened properties are correlated to the WFT by regression analysis, as presented

in the following.

5.6.1 Effects of WFT on flow spread and flow rate

By plotting the flow spread against the WFT as shown in the upper half of

Figure 5.6, it can be seen that regardless of the SFC and CSF contents, the flow

spread increased with the WFT at a gradually decreasing rate. However, at the

same WFT, a mortar with CSF added generally has a larger flow spread than a

mortar with no CSF added. To study the combined effects of the WFT and CSF

content, multi-variable regression analysis has been carried out to derive the best-

fit curves for the flow spread-WFT relation. The best-fit curves so obtained are

plotted alongside the data points and the equation and R2 value are printed in the

graph. The equation suggests that the maximum flow spread is larger when CSF is

added. The very high R2 value of 0.934 achieved with both the WFT and CSF

content considered indicates that the flow spread is governed by both the WFT

and CSF content.

By plotting the flow rate against the WFT as shown in the lower half of

Figure 5.6, it can be seen that regardless of the SFC and CSF contents, the flow

rate increased with the WFT at a more or less constant rate. However, at the same

WFT, a mortar with CSF added generally has a higher flow rate than a mortar

with no CSF added. To study the combined effects of the WFT and CSF content,

multi-variable regression analysis has been carried out to derive the best-fit curves

for the flow rate-WFT relation. As before, the best-fit curves so obtained are

plotted alongside the data points and the equation and R2 value are printed in the

graph. The two best-fit curves plotted as straight lines reveal that the flow rate-

WFT relation is basically linear. Furthermore, the very high R2 value of 0.938

achieved reveals that the WFT and CSF content are the main factors governing the

flow rate.

- 89 -

Overall, the effects of CSF are that at the same WFT, a mortar with CSF

added has larger flow spread and higher flow rate than a mortar with no CSF

added. Since the effects of CSF on the packing density and solid surface area

should have been reflected in the WFT, it seems that the CSF has certain

additional effects not reflected by the WFT. Such additional effects of CSF may

be explained as follows. First, due to their high fineness, the CSF particles would

tend to move together with the water to form a water-CSF slurry. Since the water-

CSF slurry has a larger volume than the water itself, the presence of CSF would

increase the thickness of slurry coating the cement and aggregate particles to

provide better lubrication. Second, due to their perfectly rounded shape, the CSF

particles could act as ball bearings to reduce the inter-particle friction between the

cement and aggregate particles.

5.6.2 Effects of WFT on yield stress and apparent viscosity

The yield stress is plotted against the WFT in the upper half of Figure 5.7

to illustrate the effect of WFT on yield stress. In general, the yield stress

decreased as the WFT increased. These results also reveal that the yield stress is

governed not only by the WFT, but also by the CSF content. To study the

combined effects of the WFT and CSF content, multi-variable regression analysis

has been carried out to derive the best-fit curves for the yield stress-WFT relation.

The best-fit curves so obtained are plotted alongside the data points and their

equations and R2 value are printed in the graph. It is noted that with the addition of

CSF, the best-fit curve shifts downwards and to the left. Such shifting suggests

that at the same WFT, the presence of CSF would significantly decrease the yield

stress. It is noteworthy that with both the WFT and CSF content considered in the

correlation, the R2 reaches a very high value of 0.995, indicating that the WFT and

CSF content are the controlling factors governing the yield stress.

The apparent viscosity is plotted against the WFT in the lower half of

Figure 5.7 to illustrate the effect of WFT on apparent viscosity. As for the yield

stress, the apparent viscosity also decreased as the WFT increased. Moreover, the

apparent viscosity is governed not only by the WFT, but also by the CSF content.

To study the combined effects of the WFT and CSF content, multi-variable

- 90 -

regression analysis has been carried out to derive the best-fit curves for the

apparent viscosity-WFT relation. The best-fit curves so obtained and the equation

and R2 value are presented in the graph for easy reference. With CSF added, the

best-fit curve shifts downwards and to the left. Such shifting reveals that at the

same WFT, the addition of CSF would significantly decrease the apparent

viscosity. A very high R2 value of 0.997 has been achieved in the correlation,

indicating that the apparent viscosity is dependent mainly on the WFT and CSF

content.

The above results that both the yield stress and apparent viscosity

decreased as the WFT increased are expected because a water-solid mixture with

a larger WFT should have smaller yield stress and apparent viscosity. It is more

interesting to note that the effects of CSF on the yield stress and apparent

viscosity are similar to the effects of CSF on the flow spread and flow rate. An

obvious reason is that the flow spread and flow rate are closely related to the yield

stress and apparent viscosity, respectively [Tattersall and Banfill, 1983, Kwan et

al., 2010], and thus they should be similarly affected by the CSF.

5.6.3 Effect of WFT on adhesiveness

From the adhesiveness-WFT curves plotted in Figure 5.8, it can be seen

that, compared to the W/S ratio, the WFT has a clearer effect on the adhesiveness

of mortar. On the whole, when the WFT was negative, the mortar appeared to be

rather dry and the adhesiveness was very small or zero. As the WFT increased, the

mortar became slightly wetter and the adhesiveness increased dramatically to a

certain maximum value. Then, as the WFT further increased, the mortar became

quite wet and the adhesiveness decreased. Hence, for each combination of SFC

and CSF contents, there was an optimum WFT at which the adhesiveness reached

a certain maximum value. In general, the optimum WFT giving the maximum

adhesiveness was around 0 to 0.10 µm. It may thus be said that the adhesiveness is

highest when the mortar is neither too dry nor too wet, as indicated by its WFT,

taken as a measure of wetness, falling within 0 to 0.10 µm. This is an important

guideline for the design of high-build mortar for rendering and repair, and the

- 91 -

mortar portion of high segregation stability concrete, which are required to have

high adhesiveness.

Regarding the effects of SFC and CSF contents, the test results have not

revealed any clear trend. Frankly speaking, the stepwise variations of the WFT

were not small enough to yield all the sharp peaks in the adhesiveness-WFT

curves. Hence, the maximum adhesiveness tended to be underestimated.

Nevertheless, it does seem that the addition of SFC and/or CSF is beneficial to the

adhesiveness; further tests are of course needed for confirmation.

5.6.4 Effect of WFT on cube strength

From the cube strength-WFT curves plotted in Figure 5.9, it can be seen

that the WFT has a distinct effect on the strength of mortar. When the WFT was

negative, the mortar appeared to be rather dry because the water was then not

sufficient to fill the voids. Under this situation, a certain amount of air was

entrapped in the voids causing the strength of mortar to be adversely affected.

When the WFT was positive, the water was sufficient to fill the voids. Under this

situation, as the WFT increased, the W/CM ratio also increased causing the

strength of mortar to gradually decrease. Hence, the maximum strength was

achieved at a WFT very close to 0 µm, in which case, the water was just sufficient

to fill the voids. As a general rule, therefore, to achieve the highest strength

possible, a positive but very small WFT should be adopted in the design of the

mortar mix.

5.7 Conclusions

A number of mortar samples made with triple-blended cementitious

materials containing OPC + SFC + CSF and added with different water contents

were produced for packing density, rheology, adhesiveness and strength

measurements. It was found that the addition of SFC, which is finer than OPC, to

fill the voids between OPC and the addition of CSF, which has the highest

- 92 -

fineness, to fill the voids between OPC and SFC can significantly increase the

packing density of the solid particles. However, because of the respective increase

in solid surface area, which thins down the water films coating the solid particles,

the WFT does not always increase. Relatively, the addition of SFC is less

effective in increasing the WFT but would increase the WFT over a wider range

of water content, and the addition of CSF is more effective in increasing the WFT

but would increase the WFT only at W/S ratio by volume ≤ 0.45 or W/CM ratio

by mass ≤ 0.32.

Correlation of the measured flow spread, flow rate, yield stress, apparent

viscosity and adhesiveness to the WFT revealed that the WFT plays an important

role in the fresh properties of mortar. Generally, the flow spread and flow rate

would increase and the yield stress and apparent viscosity would decrease as the

WFT increases. At the same WFT, the SFC has little effect but the CSF has great

effects on these rheological properties. This is probably because (1) the CSF,

being ultrafine, would tend to move together with the water to form a water-CSF

slurry, which has a large volume than the water itself, to provide better lubrication;

and (2) the CSF, being perfectly rounded, could act as ball bearings to reduce the

inter-particle friction between the larger solid particles. On the other hand, it is

evident that the adhesiveness is highest when the mortar is neither too dry nor too

wet, as indicated by its WFT, taken as a measure of wetness, falling within 0 to

0.10 µm.

Lastly, the strength results revealed that the WFT also plays an important

role in the hardened properties of mortar. Basically, a positive but very small

WFT is the optimum for producing the highest strength. Furthermore, the addition

of SFC and/or CSF to increase the packing density would allow the W/CM ratio

to be lowered while keeping the WFT positive to increase the strength.

- 93 -

Tables

Table 5.1 Mix proportions, packing density and WFT


mass



WFT (µm)

OPC SFC CSF FA* Water SP

M-0-0-0.6 0.193 1048 0 0 1099 205 12

0.737 (0.357)

-0.163

M-0-0-0.8 0.257 982 0 0 1030 256 11 -0.019

M-0-0-1.0 0.321 924 0 0 969 300 11 0.125

M-0-0-1.2 0.386 872 0 0 914 339 10 0.269

M-10-0-0.6 0.194 943 99 0 1098 204 13

0.749 (0.335)

-0.117

M-10-0-0.8 0.259 883 93 0 1029 255 12 0.016

M-10-0-1.0 0.323 831 87 0 968 299 12 0.149

M-10-0-1.2 0.388 784 82 0 914 338 11 0.282

M-20-0-0.6 0.195 837 198 0 1097 203 14

0.755 (0.325)

-0.094

M-20-0-0.8 0.260 784 185 0 1028 254 13 0.030

M-20-0-1.0 0.325 738 174 0 967 298 12 0.153

M-20-0-1.2 0.390 697 165 0 913 337 12 0.277

M-0-10-0.5 0.166 966 0 76 1126 167 23

0.780 (0.282)

-0.059

M-0-10-0.6 0.199 934 0 73 1088 195 22 -0.021

M-0-10-0.8 0.265 876 0 69 1020 245 21 0.057

M-0-10-1.0 0.331 824 0 65 960 290 20 0.134

M-0-10-1.2 0.397 778 0 61 907 329 19 0.211

M-10-10-0.5 0.167 858 101 76 1124 166 24

0.789 (0.267)

-0.036

M-10-10-0.6 0.200 829 98 73 1087 194 23 0.001

M-10-10-0.8 0.267 778 92 69 1019 244 22 0.075

M-10-10-1.0 0.333 732 86 65 959 289 21 0.149

M-10-10-1.2 0.400 691 82 61 906 328 20 0.222

M-20-10-0.5 0.168 750 202 76 1123 165 25

0.800 (0.250)

-0.028

M-20-10-0.6 0.201 725 196 73 1086 193 24 0.007

M-20-10-0.8 0.268 680 183 69 1018 243 23 0.078

M-20-10-1.0 0.335 640 173 65 958 288 22 0.149

M-20-10-1.2 0.402 604 163 61 905 327 20 0.220


- 94 -


Mix no. Flow

spread (mm)

Flow rate

(ml/s)

Yield stress (Pa)

Apparent viscosity

(Pas)

Adhesive- ness (g)

28-day cube

strength (MPa)

M-0-0-0.6 0.0 0.0 - - 0.0 7.0

M-0-0-0.8 5.5 0.0 - - 0.6 90.8

M-0-0-1.0 155.0 169.8 9.2 6.7 26.0 91.4

M-0-0-1.2 210.0 365.8 2.6 3.1 13.4 72.9

M-10-0-0.6 0.0 0.0 - - 0.0 84.8

M-10-0-0.8 1.5 0.0 95.8 36.2 2.7 95.8

M-10-0-1.0 195.0 202.5 3.8 2.9 19.8 95.9

M-10-0-1.2 206.0 391.3 2.8 2.6 15.9 82.1

M-20-0-0.6 0.0 0.0 - - 0.0 88.8

M-20-0-0.8 0.0 8.5 78.3 29.2 93.3 99.8

M-20-0-1.0 191.0 226.8 5.2 4.0 15.8 95.5

M-20-0-1.2 195.0 354.4 2.7 1.7 12.8 90.0

M-0-10-0.5 0.0 0.0 - - 0.0 98.6

M-0-10-0.6 0.0 0.0 - - 0.0 117.6

M-0-10-0.8 162.0 69.8 8.2 9.7 24.0 105.9

M-0-10-1.0 260.5 236.3 2.5 2.5 12.8 99.1

M-0-10-1.2 289.5 365.8 1.2 1.3 9.6 89.1

M-10-10-0.5 0.0 0.0 - - 0.0 109.1

M-10-10-0.6 2.5 0.0 - - 0.6 116.6

M-10-10-0.8 171.5 97.8 6.0 7.4 23.8 108.2

M-10-10-1.0 250.0 252 2.7 2.8 9.4 100.7

M-10-10-1.2 274.0 354.4 1.6 1.8 9.0 91.0

M-20-10-0.5 0.0 0.0 - - 0.0 87.2

M-20-10-0.6 0.0 2.9 59.2 46.9 103.6 116.0

M-20-10-0.8 191.5 155.3 5.9 5.5 26.0 112.9

M-20-10-1.0 230.5 315.0 2.6 2.9 22.6 105.9

M-20-10-1.2 246.0 365.8 2.1 2.2 13.2 96.3

- 95 -

Figures

0

20

40

60

80

100

0.1 1 10 100 1000 10000

Particle size (µm)

Per

cent

age

pass

ing

(%

) .

Figure 5.1 Particle size distributions of CSF, SFC, OPC and fine aggregate

SFC OPC CSF fine aggregate

- 96 -

-0.20

-0.10

0.00

0.10

0.20

0.30

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

WF

T (

µm)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.2 Variation of WFT with W/S ratio at different SFC and CSF contents

- 97 -

0

50

100

150

200

250

300

350

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

Flo

w s

prea

d (m

m)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

0

50

100

150

200

250

300

350

400

450

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

Flo

w r

ate

(ml/

s)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.3 Flowability versus W/S ratio

- 98 -

0

20

40

60

80

100

120

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

Yie

ld s

tres

s (P

a)SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

0

10

20

30

40

50

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

App

aren

t vi

scos

ity

(Pas

)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.4 Rheological properties versus W/S ratio

- 99 -

0

20

40

60

80

100

120

0.20 0.30 0.40 0.50 0.60

W/S ratio by volume

Adh

esiv

enes

s (g

)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.5 Adhesiveness versus W/S ratio

- 100 -

0

50

100

150

200

250

300

350

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

Flo

w s

prea

d (m

m)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Note: x1 is the WFT (µm); x2 is the CSF content (%)

0

50

100

150

200

250

300

350

400

450

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

Flo

w r

ate

(ml/

s)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%


Figure 5.6 Flowability versus WFT

R2=0.934

y=a+becx1

a=230+8x2 b=-230-8x2

c=-10

R²=0.938

y=ax1

a=1300+40x2

- 101 -

0

20

40

60

80

100

120

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

Yie

ld s

tres

s (P

a)SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%


0

5

10

15

20

25

30

35

40

45

50

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

App

aren

t vis

cosi

ty (

Pas

)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%


Figure 5.7 Rheological properties versus WFT

R2=0.995

y=a+becx1

a=1 b=135-6x2

c=-20-1.5x2

R2=0.997

y=a+becx1

a=2 b=47+x2

c=-19-1.5x2

- 102 -

0

20

40

60

80

100

120

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

Adh

esiv

enes

s (g

)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.8 Adhesiveness versus WFT

- 103 -

0

20

40

60

80

100

120

140

-0.20 -0.10 0.00 0.10 0.20 0.30

WFT (µm)

28-d

ay c

ube

stre

ngth

(M

Pa)

SFC= 0%, CSF= 0%

SFC=10%, CSF= 0%

SFC=20%, CSF= 0%

SFC= 0%, CSF=10%

SFC=10%, CSF=10%

SFC=20%, CSF=10%

Figure 5.9 Cube strength versus WFT

- 104 -

CHAPTER 6

CONCLUSIONS AND RECOMMENDATIONS

6.1 Summary of Study

An investigation was carried out on the properties of mortars with binary

and ternary blended cementitious materials. Totally three kinds of supplementary

cementitious materials (SCM), namely fly ash microsphere (FAM), superfine

cement (SFC) and condensed silica fume (CSF) were blended with ordinary

Portland cement (OPC) to produce mortar for testing. The effects of binary and

ternary blending cementitious materials with the use of these three kinds of SCMs

were studied. Based on the results, it is found that the newly developed parameter,

water film thickness, plays a key role in the performance of mortar, whereas the

interaction of the particles also imposes certain effects on the rheology of mortar.

With the WFT as an indicator, it is more effective to design the mortar tier in

concrete so that optimize the mix design of high-performance concrete (HPC)

incorporated of different kinds of SCMs. Due to the increasing sustainable

requirements, the concrete for future will be produced with high volume recycle

materials acting as cementitious materials and aggregates, making the concrete

mix design more complex and the conventional trial and error method hard to

succeed. A scientific mix design method must be developed, which should be

based on a clear and fundamental understanding of the mechanism that how these

materials influence concrete’s rheology. This study is from this perspective and

through the concepts of packing density, water film thickness and particle

interaction, this study contributes to improving our understanding of the effects of

SCM and developing a systematic mix design method.

Detailed conclusions have been given at the end of Chapters 3-5. The most

important of these can be summarized as follows:

- 105 -

6.1.1 Role of packing density

Since fresh mortar is essentially a water-solid mixture, the water content of

the mortar should impose great effect on its fresh properties. However, this effect

varies with different solid particle systems in mortar due to the water has to fill up

the voids between the solid particles first and only the water in excess of the

amount needed to fill up the voids would lubricate the solid particles. Hence, the

packing density of the solid particle system, which determines the voids content,

should be of great importance to the fresh properties of mortar too. This

importance was recognized before but for a long time, lacking of proper method

to measure packing density, especially for the fine solid particles, remains the

major obstacle. This study adopted a wet packing method, which was developed

recently in the University of Hong Kong, to evaluate the packing densities of solid

particles in different mortar mix samples. This method takes into consideration of

effects of water and SP and was operated under wet condition which is more

realistic. Compared with the traditional dry packing method, this method is less

sensitive to the compaction applied and the results determined from this method

showed a better repeatability. Due to these advantages, it is highly advocated this

wet packing method should be used, in replacement of the dry packing method, to

measure the packing density of cementitious materials or even aggregate.

6.1.2 Role of water film thickness

With the determination of packing density of mortar by the wet packing

method, the effect of packing density on the performance of mortar could be

demonstrated. The fine SCM are found effective to improve the packing density

so as the excess water content. However, the increased solid surface area brought

by the fine SCM was found to have negative effects on the flowability of mortar.

Regarding this, the author’s research team proposed a new parameter, water film

thickness (WFT), to combine the effects of packing density and solid surface area.

The WFT is determined as the excess water content to solid surface area and has

the physical meaning of the average thickness of water films coating the solid

particles. It was shown that the rheology of mortar always improved as the WFT

increased. Given the same water content, suitable adjustment of the packing of

- 106 -

solid ingredients will maximize the WFT at almost no extra cost and without

compromising other hardened properties. In the researches before, this WFT has

been demonstrated as the governing factor in cement paste and cement mortar. In

this study, the role of WFT in the mortar produced with different kinds of SCM

was further investigated.

6.1.3 Effects of fly ash microsphere

Since the filling effect of the very fine particles of fly ash microsphere

(FAM), addition of FAM up to 40% replacement content could significantly

increase the solid packing density of mortar. The addition of FAM will be more

effective to improve the WFT at a low W/S ratio, which can be also reflected as

the more significant increase in flow spread at the low W/S ratio. It is concluded

that the addition of FAM exerts its effects on rheology through the filling effect,

which increases the excess water content, the increased solid surface area, which

thins down the water films coating the solid particles, the dispersing effect, which

avoids the interlocking of angular particles and the ball bearing effect, which

reduces the friction between other solid particles.

The cube strength results reveal that the addition of FAM will improve the

strength significantly. This improvement should be resulted from the increased

packing density, rather than the pozzolanic effect which should be not obvious in

28 day. The concurrent strength-flowability performance showed that the addition

of FAM up to 40% content could increase the strength at the same flowability,

increase the flowability at the same strength, or increase both the strength and

flowability at the same time. Compared the 20% FAM content, further increase of

FAM to 40% content would not further improve the concurrent strength-

flowability performance. On the other side, 40% FAM content would not decrease

the overall performance of mortar either. Generally, FAM is considered as an

excellent SCM for improving the rheology and strength of mortar.

- 107 -

6.1.4 Effects of superfine cement

Since the filling effect of the very fine particles of the superfine cement

(SFC), addition of SFC up to 20% replacement content could significantly

increase the solid packing density of mortar. The addition of SFC would generally

increase the flowability, decreased the yield stress and apparent viscosity of

mortar mix. The addition of SFC influences the rheology through the filling effect

and the increased solid surface area. Due to the angular shape of SFC particles,

the ball bearing effect, which was found in FAM, does not exist anymore. The 28-

day cube strengths results indicated that the cube strength increased as the W/CM

ratio decreased and then decreased as the W/CM ratio decreased to beyond a

certain optimum value. The optimum W/CM ratio for maximum cube strength

was lower at a higher SFC content while the maximum cube strength was a little

bit higher at a higher SFC content.

6.1.5 Effects of condensed silica fume

Compared with FAM and SFC, the CSF was demonstrated as the most

effective one to improve the fresh and hardened properties of mortar. Firstly, due

to its ultrahigh fineness, addition of CSF could significantly increase the solid

packing density of mortar. The results determined from the wet packing method

revealed that even only 10% replacement content of CSF can drastically increase

the packing density. Secondly, addition of the ultra-fine CSF would significantly

increase the solid surface area at the same time. So high-volume addition of CSF

is not suggested and was not adopted in this study either. But at a low W/CM ratio,

the addition of CSF would largely increase the WFT so that improve the

flowability of mortar. Thirdly, due to the perfect rounded shape of CSF particles,

the ball bearing effect would reduce the friction between other solid particles

effectively. Last, due to the ultrahigh fineness, the CSF particles would tend to

move together with the water to form water-CSF slurry. Since the water-CSF

slurry has a larger volume than the water itself, the presence of CSF would

increase the thickness of the slurry coating the solid particles so that provide better

lubrication. The cube strength results also revealed that CSF could effectively

- 108 -

increase the cube strength and by now still remain the most important SCM in the

production of HPC.

6.1.6 Effects of ternary blended system

There are numerous researches about binary blended system, which is

referred to blending OPC with one kind of SCM. However, ternary blending OPC

with two kinds of SCM with different fineness would improve the particle size

distribution and further increase the packing density by the successive filling

effect. Especially adding CSF with another SCM of fineness between CSF and

OPC would be greatly beneficial to the overall performance of mortar. In this

study, both binary blended system, e.g. OPC+FAM or OPC+SFC, and ternary

blended system, e.g. OPC+FAM+CSF or OPC+SFC+CSF, were used to produce

mortar mix samples. The results showed that the ternary blended system could

more effectively benefit the packing density along with excess water content,

while not excessively increase the solid surface area. As a result, the mortars with

ternary blended cementitious materials usually show better flowability than the

ones with binary blended system. The strength results show that the ternary

blended system also appears to be more advantageous to increase the strength of

mortar. Compared with binary blended system, ternay blending CSF with FAM or

SFC would not only improve the overall strength-flowability performance, but

also allow part of CSF to be replaced by FAM or SFC to reduce the cost of

production. Considering all the advantages above, it is the author’s belief that the

ternary blended system will contract more interests with the increasing popularity

of HPC.

6.2 Conclusions

Adding SCM to improve the overall performance of concrete is no doubt a

great advancement in the development of concrete technology. However the

mechanism of how SCM affect the rheology and strength of concrete still remains

elusive, making the mix design rather difficult. The traditional strategy of

- 109 -

optimizing the mix design of cement paste/mortar/concrete is to maximize the

packing density of the particle system, which can minimize the voids content and

improve the strength significantly. However, at the same water content, packing

density is not the only parameter that governs the workability of concrete. The

solid surface area also plays an essential role. In this study, the author proposed to

take into consideration both of the two aspects in the mix design through the

newly developed parameter, water film thickness (WFT). It is the author’s belief

that it is the WFT, rather than the packing density, that should be optimized. In

other words, the mix design of the mortar fraction in concrete should be optimized

by not only controlling the water content, but also adjusting the packing density

and solid surface area. The test results in this study have demonstrated that the

WFT should be the most important parameter to be considered in the mix design

of high-performance mortar and concrete. The findings of this study have

improved the understanding of the mechanism of how SCM influence the

performance of concrete and based on these findings, a general guideline for

adopting SCM in the production of HPC is established.

To improve the workability of concrete, adding SCM finer than OPC is

beneficial only when they can increase the WFT. However, whether the WFT

would be increased is dependent on whether the proportional increase in excess

water due to the addition of SCM is larger than the proportional increase in solid

surface area. The proportional increase in excess water is in turn dependent on the

water content while the proportional increase in solid surface area is independent

on the water content. When SCM are added in a mortar with a lower water content,

the increase in excess water will be higher compared with the situation with a

higher water content. That is the reason that addition of SCM is usually more

effective in increasing the WFT at a relative low W/S ratio. When CSF is blended

together with another SCM with a mean particle size of several micrometers, the

packing density will be effectively increased so as the amount of excess water,

while the corresponding increase in solid surface area will not be too excessive

compared with blending CSF alone. As a result, ternary blended system is more

effective in increasing the WFT in the mix design. Besides, the effects of particle

interaction should be considered at the same time. Adding some SCM with

spherical particle shape will be beneficial for the concrete to achieve higher

- 110 -

flowability at the same WFT value through the ball bearing effect. Furthermore,

some ultrafine particles of SCM tend to move with water to form slurry between

the other particles to provide better lubrication. So, taking the advantage of the

particle interaction should be an effective strategy especially when the WFT has

been already adjusted in the mix design. To ensure the minimum flowability, the

WFT is always suggested to be set at a positive value.

To improve the adhesiveness of mortar, the WFT should be set at a value

between 0 to 0.15 µm. It is observed that the mortar mix usually achieve their

maximum adhesiveness at different water content. The WFT could be used as a

more effective indicator of the optimum wetness for the mortar mix compared

with the W/S ratio. So far, there has been little research on adhesiveness since the

lacking of scandalized measurement method. But increasing the adhesiveness of

mortar should be an effective method to produce high-build mortar for rendering

and concrete repair, and the mortar portion of concrete requiring high

cohesiveness (which is highly dependent on the ability of the mortar to adhere to

the aggregate particles).

To improve the strength of concrete, addition of SCM finer than OPC will

be preferred in the mix design since it can effectively increase the packing density

so that reduce the water content required in a concrete. With the lowering of the

W/CM ratio, the strength will be dramatically improved. Incorporating SCM with

pozzolanic activity will further increase the strength through the pozzolanic

reaction which can improve the microstructure of the concrete. Usually, the

concrete mixed with fine SCM appears very cohesive, leading to less bleeding

water and thus stronger bonding force of the interfacial transition zone which

finally dominates the strength of concrete. Ternary blending OPC with two kinds

of SCM with different levels of fineness will achieve a better packing density, as a

result, higher strength, compared with binary blending OPC with one kind of

SCM. To achieve the peak strength, a WFT value slightly above zero is suggested

in the mix design.

Summing up, SCM can be of great effectiveness in improving the overall

performance of concrete if properly added. Based on the theories of packing of

- 111 -

solid particles, WFT and particle interaction, we can improve the understanding of

the roles of SCM in concrete and at last make our mix design more scientific and

the concrete for tomorrow perform better.

6.3 Recommendations for Further Work

This thesis has raised the concepts of WFT and presented a comprehensive

study on its role in the performance of mortar. The following areas are

recommended for future work:

(1) The particle interaction between the SCM and other solid particles is very

complex which would also have great impacts on the rheology of mortar.

This study has discussed these effects preliminarily. Further research,

including computer modelling, is suggested.

(2) The concept of WFT should be extended to concrete tier as the same, since

the role of WFT has not been studied and demonstrated in the concrete tier

so far. Hence, research on this topic is highly recommended.

(3) Similar to the excess water, acting as the water film coating the solid

particles, the excess paste and excess mortar would also act as paste film

and mortar film coating the aggregates. The relative paste film thickness

and mortar film thickness are considered as important parameters to

influence concrete’s performance. The author’s postgraduate study could

be considered as one part of this systematic study and further investigation

on the combined effects of WFT, paste film thickness and mortar film

thickness on performance of concrete is highly recommended.

- 112 -

REFERENCES

ACI Committee 212, “Guide for the Use of High-Range Water-Reducing

Admixtures (Superplasticizers) in Concrete”, ACI 212.4R-04, American

Concrete Institute, USA, 2004, 13 pp.

ACI Committee 226, “Use of fly ash in concrete”, ACI 226.3R-87, ACI Materials

Journal, Vol. 84, No.5, 1987, pp 381-408.

Aїtcin, P. -C., “Cements of yesterday and today: Concrete of tomorrow”, Cement

and Concrete Research, Vol. 30, No. 9, 2000, pp 1349-1359.

Aїtcin, P. -C., “Improving the sustainability of high performance concrete”,

Proceeding of 9th International Symposium on High Performance Concrete,

Rotorua, New Zealand, 2011, keynote paper.

Aїtcin, P. -C., High-Performance Concrete, E & FN Spon, London and New York,

1998, 591 pp.

Alves, M., Cremonini, R. and Dal Molin, D., "A comparison of mix proportioning

methods for high-strength concrete", Cement and Concrete Composites, Vol.

26, No.6, 2004, pp 613-621.

Artelt, C. and Garcia, E., “Impact of superplasticizer concentration and of ultra-

fine particles on the rheological behaviour of dense mortar suspensions”,

Cement and Concrete Research, Vol. 38, No. 5, 2008, pp 633-642.

Bharatkumar, B. H., Narayanan, R., Raghuprasad, B. K. and Ramachandramurthy,

D. S., “Mix proportioning of high performance concrete”, Cement and

Concrete Composites, Vol. 23, No. 1, 2001, pp 71-80.

Claisse, P. A., Lorimer, P. and Al Omari, M., “Workability of cement pastes”,

ACI Materials Journal, Vol. 98, No. 6, 2001, pp 476-482.

Collepardi, M. and Valente, M., “Recent developments in superplasticizers”, ACI

SP-239, 2006, pp 1-14.

Collins, F. and Sanjayan, J. G., “Effects of ultra-fine materials on workability and

strength of concrete containing alkali-activated slag as the binder”, Cement

and Concrete Research, Vol. 29, No. 3, 1999, pp 459-462.

De Larrard, F., Concrete Mixture Proportioning: A Scientific Approach, E&FN

Spon, New York, 1999, Chapter 2, pp 77-221.

De Larrard, F. and Sedran, T., “Optimization of ultra-high-performance concrete

by the use of a packing model”, Cement and Concrete Research, Vol. 24, No.

- 113 -

6, 1994, pp 997-1009.

Domone, P. and Soutsos, M., “An approach to the proportioning of high-strength

concrete mixes”, Concrete International-Design and Construction, Vol. 16,

No.10, 1994, pp 26-31.

Duval, R. and Kadri, E. H., “Influence of silica fume on the workability and the

compressive strength of high-performance concretes”, Cement and Concrete

Research, 1998, Vol. 28, No. 4, pp 533-547.

Erdem, T. K. and Kırca, Ö., “Use of binary and ternary blends in high strength

concrete”, Construction and Building Materials, Vol. 22, No. 7, 2008, pp

1477-1483.

Feng, N. Q., Shi, Y. X. and Hao, T. Y., “Influence of ultrafine powder on the

fluidity and strength of cement paste”, Advances in Cement Research, Vol.

12, No. 3, 2000, pp 89-95.

Ferraris, C. F., Obla, K. H. and Hill, R., “The influence of mineral admixtures on

the rheology of cement paste and concrete”, Cement and Concrete Research,

Vol. 31, No. 2, 2001, pp 245-255.

FIP Commission on Concrete, FIP State-of-the-Art Report: Condensed Silica

Fume in Concrete, Thomas Telford, London, 1988, 37 pp.

Fung, W. W. S. and Kwan, A. K. H., “Role of water film thickness in rheology of

CSF mortar”, Cement and Concrete Composites, Vol. 32, No. 4, 2010, pp

255-264.

Fung, W. W. S., Kwan, A. K. H. and Wong, H. H. C., “Wet packing of crushed

rock fine aggregate”, Materials and Structures, Vol. 42, No. 5, 2009, pp 631-

643.

Gao, P. W., Deng, M. and Feng, N. Q., “The influence of superplasticizer and

superfine mineral powder on the flexibility, strength and durability of HPC”,


Glavind, M. and Munch-Petersen, C., “Green concrete - a life cycle approach”,

Sustainable Concrete Construction, MPG Books, Bodmin, UK, 2002, pp

771-786.

Hattori, K., “Experience with Mighty superplasticizer in Japan”, ACI SP-62, 1978,

pp 37-66.

Helmuth R. A., “Structure and rheology of fresh cement paste”, Proceedings of

7th International Congress of Chemistry of Cement, Paris, 1980, pp 16-30.

- 114 -

Hunger, M. and Brouwers, H., “Flow analysis of water-powder mixtures:

Application to specific surface area and shape factor”, Cement and Concrete

Composites, Vol. 31, No. 1, 2009, pp 39-59.

Isaia, G., Gastaldini, A. and Moraes, R., “Physical and pozzolanic action of

mineral additions on the mechanical strength of high-performance concrete”,

Cement and Concrete Composites, Vol. 25, No. 1, 2003, pp 69-76.

Kaci, A., Bouras, R., Phan, V.T., Andréani, P.A., Chaouche, M. and Brossas, H.,

“Adhesive and rheological properties of fresh fibre-reinforced mortars”,

Cement Concrete Composite, Vol. 33, No. 2, 2011, pp 219-224.

Khatri, R. P., Sirivivatnanon, V. and Gross, W., “Effect of different

supplementary cementitious materials on mechanical properties of high

performance concrete”, Cement and Concrete Research, Vol. 25, No. 1, 1995,

pp 209-220.

Kohno, K. and Komatsu, H., “Use of ground bottom ash and silica fume in mortar

and concrete”, Proceedings of ACI Second International Conference on Fly

Ash, Silica Fume, Slag, and Natural Pozzolans in Concrete, Madrid, Spain,

1986, pp 1279-1292.

Kwan, A. K. H., “Development of high performance concrete for Hong Kong”,

Proceedings of Materials Science and Technology in Engineering

Conference, Hong Kong, 2003, pp 1-16.

Kwan, A. K. H., “Use of condensed silica fume for making high-strength, self-

consolidating concrete”, Canadian Journal of Civil Engineering, Vol. 27, No.

4, 2000, pp 620-627.

Kwan, A. K. H., Cai, Y. B. and Chan, H. C., “Development of very high strength

concrete for Hong Kong”, HKIE Transactions, 1995, Vol. 2, No. 2, pp 9-16.

Kwan, A. K. H., Chen, J. J. and Fung, W. W. S., “Effects of superplasticizer on

rheology and cohesiveness of CSF cement paste”, Advances in Cement

Research, 2012, (in press).

Kwan, A. K. H. and Fung, W. W. S., “Effects of CSF content on rheology and

cohesiveness of mortar”, Magazine of Concrete Research, Vol. 63, No. 2,

2011, pp 99-110.

Kwan, A. K. H. and Fung, W. W. S., “Packing density measurement and

modelling of fine aggregate and mortar”, Cement and Concrete Composites,

Vol. 31, No. 6, 2009, pp 349-357.

- 115 -

Kwan, A. K. H., Fung, W. W. S. and Wong, H. H. C., “Water film thickness,

flowability and rheology of cement-sand mortar”, Advances in Cement

Research, Vol. 22, No. 1, 2010, pp 3-14.

Kwan, A. K. H. and Ng, I. Y. T., “Improving performance and robustness of SCC

by adding supplementary cementitious materials”, Construction and Building

Materials, Vol. 24, No. 11, 2010, pp 2260-2266.

Kwan, A. K. H. and Wong, H. H. C., “Effects of packing density, excess water

and solid surface area on flowability of cement paste”, Advances in Cement

Research, Vol. 20, No. 1, 2008a, pp 1-11.

Kwan, A. K. H. and Wong, H. H. C., “Packing density of cementitious materials:

part 2 - packing and flow of OPC + PFA + CSF”, Material and Structures,

Vol. 41, No. 4, 2008b, pp 773-784.

Lee, S. H., Kim, H. J., Sakai, E. and Daimon, M., “Effect of particle size

distribution of fly ash-cement system on the fluidity of cement pastes”,


Li, L. G. and Kwan, A. K. H., “Mortar design based on water film thickness”,

Construction and Building Materials, Vol. 25, No. 5, 2011, pp 2381-2390.

Lothenbach, B., Scrivener, K. and Hooton, R. D., “Supplementary cementitious

materials”, Cement and Concrete Research, Vol. 41, No. 12, 2011, pp 1244-

1256.

MacArthur, J. M., Read, A. S. and Cole, R. N., “Grade 100 concrete in Hong

Kong”, Asia Engineer, September, 1996, pp 9-11.

Malhotra, V. M., “Making concrete greener with fly ash”, Concrete International,

Vol. 21, No. 5, 1999, pp 61-66.

Malhotra, V., “Role of supplementary cementing materials in reducing

greenhouse gas emissions”, Concrete technology for a sustainable

development in the 21st century, 2000, pp 226-235.

Mehta, P. K., “Reducing the Environmental Impact of Concrete”, Concrete

International, Vol. 23, No. 10, 2001, pp 61-66.

Mehta, P. K., “Concrete technology for sustainable development”, Concrete

International, Vol. 21, No. 11, 1999, pp 47-53.

Mehta, P. K. and Gjørv, O., “Properties of Portland cement concrete containing

fly ash and condensed silica-fume”, Cement and Concrete Research, Vol. 12,

No. 5, 1982, pp 587-595.

- 116 -

Meyer, A., “Experiences in the Use of Superplasticizers in Germany”,

Superplasticizers in concrete, ACI SP-62, 1979, pp 21-36.

Meyer, C., “The greening of the concrete industry”, Cement and Concrete

Composites, Vol. 31, No. 8, 2009, pp 601-605.

Nanthagopalan, P., Haist, M., Santhanam, M. and Müller, H. S., “Investigation on

the influence of granular packing on the flow properties of cementitious

suspensions”, Cement and Concrete Composites, Vol. 30, No. 9, 2008, pp

763-768.

Neville, A. M., Properties of Concrete, Fourth Edition, Prentice Hall, Harlow, UK,

1995.

Okamura, H. and Ouchi, M., “Self-compacting concrete”, Journal of Advanced

Concrete Technology, Vol. 1, No. 1, 2003, pp 5-15.

Papadakis, V. G., “Effect of supplementary cementing materials on concrete

resistance against carbonation and chloride ingress”, Cement and Concrete


Parrott, L. J., “The selection of constituents and proportions for producing

workable concrete with a compressive cube strength of 80 to 110 N/mm2”,

Technical Report TRA 416, May, 1969, Cement and Concrete Association,

London.

Powers, T. C., The Properties of Fresh Concrete, John Wiley & Sons, New York,

USA, 1968, 664 pp.

Quercia, G., Hüsken, G. and Brouwers, H., “Water demand of amorphous nano

silica and its impact on the workability of cement paste”, Cement and

Concrete Research, Vol. 42, No. 2, 2012, pp 344-357.

Read, T. and MacArthur, J. M., “High strength concrete - an approach to its

development in Hong Kong”, Proceedings of Hong Kong Concrete

Technology and Construction Conference, Hong Kong Institution of

Engineers, November, 1990, pp 73-99.

Rizwan, S. A. and Bier, T. A., “Self-consolidating mortars using various

secondary raw materials”, ACI Materials Journal, Vol. 106, No. 1, 2009, pp

25-32.

Sedran, T., de Larrard, F., Hourst, F. and Contamines, C., “Mix design of self-

compacting concrete”, Proceedings of the International RILEM Conference

on Production Methods and Workability of Concrete, Paisley, Scotland, 1996,

- 117 -

pp 439-450.

Svarovsky, L. and Board, B. M. H., Powder testing guide: methods of measuring

the physical properties of bulk powders, Elsevier Applied Science Publishers

Ltd, England, 1987.

Tattersall, G. H. and Banfill, P. F. G., The Rheology of Fresh Concrete, Pitman,

Boston, 1983.

Wang, A. Q., Zhang, C. Z. and Zhang, N. S., “Study of the influence of the

particle size distribution on the properties of cement”, Cement and Concrete


Wong, H. H. C. and Kwan, A. K. H., “Rheology of cement paste: Role of excess

water to solid surface area ratio”, Journal of Materials in Civil Engineering,

Vol. 20, No. 2, 2008a, pp 189-197.

Wong, H. H. C. and Kwan, A. K. H., “Packing density of cementitious materials:

part 1 - measurement using a wet packing method”, Materials and Structures,

Vol. 41, No. 4, 2008b, pp 689-701.

Wong, H. H. C. and Kwan, A. K. H., Packing density of cementitious materials:

measurement and modelling, Magazine of Concrete Research, Vol. 60, No. 3,

2008c, pp 165-175.

Yahia, A., Tanimura, M. and Shimoyama, Y., “Rheological properties of highly

flowable mortar containing limestone filler-effect of powder content and

W/C ratio”, Cement and Concrete Research, Vol. 35, No. 3, 2005, pp 532-

539.

Yu, A. B., Bridgwater, J. and Burbidge, A., “On the modelling of the packing of

fine particles”, Powder Technology, Vol. 92, No. 3, 1997, pp 185-194.

Zhang, C., Wang, A., Tang, M. and Liu, X., The filling role of pozzolanic material,

Cement and Concrete Research , 1996, Vol. 26, No. 6, pp 943-947.

Zhang, X. and Han, J., “The effect of ultra-fine admixture on the rheological

property of cement paste”, Cement and Concrete Research, Vol. 30, No. 5,

2000, pp 827-830.

Zheng, H. W., Chan, F. W. Y. and Kwan, A. K. H., “High-performance concrete

for green construction”, Proceedings of 7th International Conference on Tall

Buildings, Hong kong, China, 2009, pp 163-170.

- 118 -

LIST OF PUBLICATIONS DURING CANDIDATURE

Refereed Journal Publications

Li, Y., Chen, J. J. and Kwan, A. K. H., “Role of water film thickness in fresh and

hardened properties of mortar”, Advances in Cement Research, (accepted).

Kwan, A. K. H. and Li, Y., “Effects of fly ash microsphere on rheology,

adhesiveness and strength of mortar”, (under preparation)

Li, Y. and Kwan, A. K. H., “Ternary blending of cement with fly ash microsphere

and condensed silica fume to improve the performance of mortar”, (under

preparation)

Conference Publications

Chen, J. J., Li, Y. and Kwan, A. K. H., “Addition of chemical and mineral

admixtures to improve packing density of powder content in high-

performance concrete”, Proceeding of 9th International Symposium on High

Performance Concrete - Design, Verification & Utilization, Rotorua, New

Zealand, August, 2011.

Li, Y., Chen, J. J. and Kwan, A. K. H., “Role of water film thickness in mix

design of high-performance concrete”, Proceeding of the 2nd International

Conference on Microstructure Related Durability of Cementitious

Composites, Amsterdam, Netherlands, April, 2012.

fulltext (22)

Documents