ORIGINAL ARTICLE

Effect of silica fume, steel fiber and ITZ on the strengthand fracture behavior of mortar

Xiao Hui Wang Æ Stefan Jacobsen ÆSiaw Foon Lee Æ Jian Ying He Æ Zhi Liang Zhang

Received: 25 September 2008 / Accepted: 22 January 2009 / Published online: 30 January 2009

� RILEM 2009

Abstract Two sets of parameters known to affect

the quality and thickness of the interfacial transition

zone (ITZ), i.e. water/binder ratio and content of silica

fume were varied in a series of mortars without and

with steel fiber. Compressive and three-point bending

tests were performed and the dissipated energies were

calculated. Nanoindentation characteristics of the

steel fiber–matrix and fiber–matrix-aggregate interfa-

cial zones in the steel fiber reinforced mortars were

studied. Influence of water/binder ratio, steel fiber,

silica fume and ITZ on the strength and toughness of

the mortar was analyzed, respectively. It is found that

mortar compressive strength can be increased with

low volume addition of steel fiber if the air content is

well controlled; the interfacial characteristic and

microstructural morphology near the fiber surface

play a critical role on the three-point bending strength

and the toughness of the steel fiber reinforced mortar.

Keywords Steel fiber � Silica fume �Strength and toughness � Nanoindentation �Interfacial transition zone (ITZ)

1 Introduction

Concrete is a highly complex and heterogeneous

composite material. The properties of the concrete

depend on the aggregate, matrix, steel reinforcement

and/or fiber, as well as the interfacial transition zone

(ITZ) between the aggregate and matrix, and the steel

and/or fiber and matrix. Those transition zones

determine many of the important properties of

concrete, such as strength, cracking and fracture

behavior. In studies [1–3] carried out to determine the

influence of aggregate, matrix and ITZ properties on

the tensile and compressive strengths of concrete, it

was found that the interfacial bond was the deciding

factor for the tensile strength and played a little role

on the compressive strength. It was concluded that

the properties of the ITZ may have a moderate

influence on mechanical properties of concrete and

have a drastic effect on the mechanical properties of

fiber reinforced cement composites [4].

In addition, the ITZ properties have a particular

importance on the cracking mechanism and fracture

behavior of concrete [3, 5, 6]. Fracture of concrete

was shown to be influenced by the presence of the

paste–aggregate interface [3, 6]. The strength of the

X. H. Wang (&)

Department of Civil Engineering, Shanghai Jiaotong

University, Shanghai 200240, China

e-mail: w_xiaoh@163.com

X. H. Wang � S. Jacobsen � S. F. Lee �J. Y. He � Z. L. Zhang

Department of Structural Engineering, Faculty

of Engineering Science and Technology, Norwegian

University of Science and Technology (NTNU),

7491 Trondheim, Norway

Materials and Structures (2010) 43:125–139

DOI 10.1617/s11527-009-9475-1

interface affected the fracture energy in different

ways depending on the shape of the particles; the

critical crack opening decreased when the interface

was strong, approaching the value corresponding to

the matrix, and it increased when the crack was rough

[6]. As the w/c ratio increased, the porosity in ITZ

was also increased, resulting in the initiation and

development of cracks in this zone [7].

The above mentioned properties and behaviors of

cementitious materials at the macroscale are all

significantly affected, if not dominated, by their

structural features and properties at the micro/nano-

scale where the deterioration and failure process

starts—properties at ITZ. The morphological charac-

teristic of the microstructure of ITZ was primarily

characterized via electron microscopy. Throughout

those studies, Scanning Electron Microscopy (SEM)

[3, 8], Environmental Scanning Electron Microscopy

(ESEM) [8] and Transmission Electron Microscopy

(TEM) [9] have been extensively used. In order to

investigate and quantify microstructural gradients

across the ITZ, backscattered electron imaging [10,

11] has been also used. However, for the above

mentioned microscopical techniques, only two

dimensional sections of a three-dimensional micro-

structure can be observed [11].

It is not until recent years that progress and

improved availability of advanced instruments and

techniques, such as depth-sensing micro/nanoinden-

tation, have made it possible to study mechanical

properties of various micro/nanoscale features in

cement-based materials. Microindentation testing

was firstly used to characterize gradients in mechan-

ical properties to determine the mechanical

characteristics of microsize zones in various cement

based composites [12–14]. Due to the limitations of

microindentation, such as larger indent comparing

with the thickness of ITZ [15], nanoindentation,

which has already been used for many other

materials [16], began to be used in the cementitious

composites. By using the nanoindentation tech-

nique, better representation of heterogeneous fea-

tures exhibited by cementitious materials can be

obtained [17]. Till now, nanoindentation was widely

used to measure elastic modulus and hardness of

cement paste cured at different conditions [18–20].

Few research works were focused on the studying

of the ITZ between a rigid inclusion and matrix

[12, 21, 22].

In the present paper, in order to link the nano-

mechanical properties of the ITZ with the macro-

scopic mechanical properties, mortars without and

with steel fiber were prepared; where water/binder

ratios of 0.3 and 0.5, and 0% and 10% contents of

silica fume were considered. Uniaxial compressive

strength testing and three-point bending strength

testing were performed. Dissipated energies Welastic

and Wpost-elastic before and after maximum elastic load

were calculated to analyze the effect of steel fiber on

the toughness of the mortar. Samples prepared from

the same steel fiber reinforced mortars were used for

nanoindentation test and the steel fiber–matrix and

fiber–matrix-aggregate interfaces were studied. The

whole sizes of each indented areas were determined

by Scanning Electron Microscopy (SEM) and bond

conditions across the interfaces were observed.

Characteristics of the profiles of the interfacial

transition zones were studied. It is found that the

compressive strength of the mortar can be increased

due to the low volume (0.3 vol%) addition of steel

fiber on condition that the air content of the hardened

mortar is well controlled; while the three-point

bending strength and the toughness of the steel fiber

reinforced mortar are mainly dependent on the

interfacial characteristics and microstructural mor-

phology near the fiber surface, which are mainly

related to the water/binder ratio, the presence of silica

fume, ITZ thickness as well as voids and microcracks

within those zones.

2 Experimental programme

2.1 Materials

Nine types of mortars were prepared. The detailed

information of the specimens can be seen in Table 1,

where specimens were identified with numbers

designation: the first two numbers indicating different

water/binder ratio, the second two numbers corre-

sponding to the different content of silica fume, the

last two numbers corresponding to the different

content of steel fiber. As an example, specimen

031003 implies a mortar with water/binder ratio 0.3,

having 10% silica fume and 0.3 vol% steel fiber.

The mortar mixtures were prepared with cement,

sands 0–4 and 0–2 mm, water, limestone powder,

and/or silica fume, and/or steel fiber. Norcem Anlegg

126 Materials and Structures (2010) 43:125–139

cement which is an ordinary Portland cement was

used. Silica fume (Elkem Microsilica grade 940-U) in

powder form with the content of SiO2 larger than

90% was used. The mass retained on 45 micron sieve

is small than 1.5%. The limestone powder, used as a

filler, has a fineness of 57.2%\24 lm, with a specific

weight of 2,730 kg/m3. Sands 0–4 and 0–2 mm have

fineness modulus of 2.37 and 1.51, respectively. The

percentages of particle \0.125 mm of sands 0–4 and

0–2 mm are 6.1% and 23.6%, respectively. Glenium

151 was incorporated in all mixtures as a superplast-

icizer, having 15% solids content and a density of

1,030 kg/m3. Straight high carbon steel fibre OL13/.16

has a length of 13 mm and a diameter of 0.16 mm,

and a tensile strength more than 2000 MPa.

2.2 Proportions of mixes used

The matrix volume of all mixes is 410 l/m3. The

characteristics of the mortar mixes are shown in

Table 2, where the numbers in the brackets of column

4 and 5 indicates the contents of the limestone and

Glenium 151 as percentages of the weight of cement,

respectively.

2.3 Mixing procedures

The mortars were mixed in a flat-bottomed mixer

with maximum volume of 12 l. For the mortars

without steel fiber, the mixing procedure was as

follows: (1) cementitious materials (including silica

fume) and sands were blended for 1 min at lower

speed; (2) following the next 4 min, half of the

mixing water was firstly added during the mixing,

then all the superplasticizer and the remaining water

were added; (3) stop the mixing, substances sticking

in the sides of the mixer were cleaned and mixed into

the mixes within 1 min; (4) the mixture was allowed

to rest for 5 min, then mixed for an additional 1 min

at lower speed. For the steel fiber mortar, after

finishing the adding of the water and superplasticizer

in Step 2, the steel fibers were added. It is suggested

Table 1 The detailed

information of the

specimens

Name Water/binder ratio Silica fume (percentage

of the weight of cement)

Steel fiber

(volume percentage)

030000 0.3 0.0 0.0

030003 0.0 0.3

030010 0.0 1.0

031000 10.0 0.0

031003 10.0 0.3

050000 0.5 0.0 0.0

050003 0.0 0.3

051000 10.0 0.0

051003 10.0 0.3

Table 2 Characteristics of the mortar mixes

Name Mix proportion (kg/m3)

Anlegg cement Free water Limestone Glenium 151 Sand 0–4 mm Sand 0–2 mm Silica fume Steel fiber

030000 542.5 153.1 47.2 (8.7%) 11.39 (2.1%) 1,406.8 248.3 0.0 0.0

030003 543.0 153.7 47.2 (8.7%) 10.86 (2.0%) 1,399.4 246.9 0.0 24.2

030010 543.9 153.5 47.3 (8.7%) 11.42 (2.1%) 1,382.1 243.9 0.0 78.0

031000 483.4 149.2 47.4 (9.8%) 12.09 (2.5%) 1,406.8 248.3 48.3 0.0

031003 483.8 149.4 47.4 (9.8%) 12.10 (2.5%) 1,399.4 246.9 48.4 23.4

050000 411.2 201.4 47.3 (11.5%) 4.93 (1.2%) 1,406.8 248.3 0.0 0.0

050003 411.5 201.5 47.3 (11.5%) 4.94 (1.2%) 1,399.4 246.9 0.0 23.4

051000 367.9 196.9 47.8 (13.0%) 6.36 (1.73%) 1,406.8 248.3 36.8 0.0

051003 368.2 197.1 47.9 (13.0%) 6.37 (1.73%) 1,399.4 246.9 36.8 23.4

Materials and Structures (2010) 43:125–139 127

that the fibers should be added in small batches to get

good dispersion of fibers in the mixture.

2.4 Rheology and curing

For each type of fresh mortar, slump test with a small

cone and rheological measurements with the ConTec

Viscometer were performed. Then, 50 9 50 9

50 mm cubes were cast for the compressive test and

40 9 40 9 160 mm prisms were cast for three-point

bending strength test and nanoindentation test. During

the casting, steel molds were vibrated in order to

evacuate parts of the entrapped air. Then, surfaces of

the specimens were carefully smoothed and covered

with plastic sheets. Specimens were demolded after

24 h, and then each specimen was weighed in air and

water, respectively. After that, the specimens were

cured at 20�C under water for 28 days.

2.5 Sample preparation for studying interfacial

properties

After 28 days curing in a water bath, the central part

of the 40 9 40 9 160 mm prism was taken out using

a diamond saw. This part was cut into small samples

of the size about 15 9 15 9 15 mm, which were

embedded in epoxy resin in air and prepared for

grinding and polishing.

For fiber-reinforced composites, due to the lack of

published grinding and polishing procedures in the

literature, the best possible grinding and polishing

procedures were finally determined after several trial

tests. The TegraForce-5 grinding and polishing

machine was used. The MD-Piano plates (220, 600

and 1200) were selected for the coarse grinding to

obtain flat surface. The Largo (9 lm), Dur (3 lm)

and Dac (1 lm) were selected for the fine polishing.

The Nap was used for the final polishing. At every

step, a microscope was used to check the effective-

ness of the grinding and polishing and ultrasonic bath

cleaning was performed to remove all dust and

diamond particles.

2.6 Nano-mechanical properties of ITZ

In order to evaluate the effect of nano-mechanical

properties of the ITZ on the macroscopic mechanical

properties of mortar, such as strength, cracking and

toughness, a Triboindenter was used to study the ITZ

properties. A typical representation of the indentation

load versus indentation depth or displacement for an

indentation experiment is shown in Fig. 1. The elastic

modulus E and hardness H of the sample was

calculated as follows [16]

S ¼ dP

dh¼ 2

ffiffiffi

pp Er

ffiffiffi

Ap

ð1Þ

E ¼ ð1� m2Þ � 1

Er

� ð1� m2i Þ

Ei

� ��1

ð2Þ

H ¼ Pmax

Að3Þ

where P is the indentation load; h is the indentation

depth; S is the initial unloading stiffness; Pmax is the

peak indentation load; A is the is the projected area of

the elastic contact; Er is the reduced elastic modulus;

E and m are Young’s modulus and Poisson’s ratio for

the specimen and Ei and mi are the same parameters

for the indenter. For the indenter used the present

experiments, the elastic modulus Ei = 1,140 GPa and

the Poisson’s ratio mi = 0.07.

For the microhardness profiles of the ITZ, it was

reported that in the vicinity of the inclusion surface

there is a gradient in the microhardness, but in the

bulk paste it becomes relatively constant. The trends

in the gradients can be classified into four types as

shown in Fig. 2 [13]. For the type I curve, there is an

Fig. 1 A typical representation of the indentation load versus

indentation depth [16]

128 Materials and Structures (2010) 43:125–139

increase in the microhardness of the matrix in the

vicinity of the inclusion. The reason may lie in (1)

the inclusion and the matrix are well bonded at the

interface; (2) the near surface ITZ is rich in massive

CH [13]. Deviations from these conditions can lead to

changes in the shape of the curve and can account for

shapes such as II, III, and IV in Fig. 2.

3 Results and discussion

3.1 Properties of fresh and hardened mortar

Properties of the fresh and hardened mortar are

shown in Table 3. In order to assess workability and

consistency of the fresh mortar, density of the fresh

mortar, air content, slump, flow diameter, yield value

and plastic viscosity were measured, respectively.

Just as summarized by Banfill [23], with the increase

of the water/binder ratio, the yield stress and plastic

viscosity decrease, indicating a better workability

and consistency for mortars with large water/binder

ratio. In addition, due to the addition of the silica

fume, the yield stress increases for mortars with low

water/binder ratio, indicating a workability loss due

to the addition of the silica fume. The workability

and consistency is obviously influenced by the

addition of steel fiber in mortars with w/b = 0.3.

For instance, with the increased addition of steel

fiber from 0.0 to 1.0 vol%, yield stress and plastic

viscosity increase greatly. For specimen 030010, it is

too stiff to be measured in the ConTec Viscometer.

However, for mortars with w/b = 0.5, the workabi-

lity and consistency is less influenced by the addition

of low content of steel fiber. The yield stress and

plastic viscosity of specimen 050003 are even

smaller than those of specimen 050000, mainly due

to the low content of steel fiber and partly insuffi-

cient dispersion of fibers.

The air content of fresh mortars with w/b = 0.3

increases slightly with the addition of silica fume.

Along with this effect is a decrease in density after

demoulding. However, due to the use of superplast-

icizer in the mortars with high w/c ratios (e.g.,

w/b = 0.5), segregation of the mixture may occur [24]

and the air content of fresh mortars with silica fume is

smaller than that of mortars without silica fume.

Fig. 2 Classification of microhardness profile in the interfacial

transition zone (ITZ) around a rigid inclusion in a cement paste

matrix [13]

Table 3 Properties of the fresh and hardened mortar

Name Fresh mortar Hardened mortar

Slump

(mm)

Flow diameter

(mm)

Yield value

(Pa)

Plastic viscosity

(Pa�s)

Air content

(%)

qtheor(voidfree) qdemould k(%)

030000 101 240/265 8 30.2 1.9 2,458 2,359 (7.62) 4.03

030003 94 180/205 11 43.0 2.3 2,473 2,358 (7.95) 4.67

030010 60 – – – 3.2 2,510 2,399 (19.26) 4.43

031000 105 230/235 30 14.5 4.0 2,445 2,323 (10.75) 4.99

031003 95 200/205 42 15.8 2.9 2,460 2,324 (7.85) 5.54

050000 100 265/265 8 6.5 3.4 2,367 2,293 (4.04) 3.14

050003 98 260/260 7 5.7 4.0 2,383 2,302 (3.58) 3.39

051000 110 230/245 10 2.8 2.5 2,358 2,244 (9.66) 4.84

051003 103 245/245 10 5.1 1.5 2,373 2,291 (4.92) 3.47

Note: Number in brackets is the standard deviation

Materials and Structures (2010) 43:125–139 129

In Table 3, the air content of the hardened mortar

k was calculated as follows [25]

kð%Þ ¼ 1� qdemould

qtheorðvoidfreeÞð4Þ

where qtheor(voidfree) is the theoretical void free density

of the mortar and qdemould is the measured density

after demoulding. It seems that the air content of the

hardened mortar does not agree well with the air

content measured on the fresh mortar. This may result

from the vibration of the molds during the casting. It

also can be seen from Table 3 that the air content of

the hardened mortars with both water/binder ratios is

increased by the addition of the silica fume.

3.2 Compressive and bending strengths

Uniaxial compressive strength testing of the mortars

was performed on 50 mm cubes after 28 days of

curing. Test results are shown in Table 4. Each result

presented the average of six companion specimens.

Three-point bending strength testing was performed

on 40 9 40 9 160 mm prisms after 28 days of

curing, where the support span is 100 mm. The

bending strength presented is the average of three

companion specimens.

It can be seen from Table 4 that the compressive

and bending strengths of the mortars are influenced by

the addition of silica fume and steel fiber. For the

mortars without steel fiber, both the compressive

strength and bending strength are increased by the

addition of silica fume, where the increase of

compressive strength is very obvious for mortars with

low and high water/binder ratio, as pointed out by

Kjellsen et al. [26]. For the steel fiber reinforced

mortar, the compressive strength is also increased by

the addition of silica fume. However, the bending

strength is little influenced or even reduced. For

mortars without and with silica fume, although it was

reported that the fibers are increasingly being used to

enhance the strength and toughness of the cementi-

tious composites [27], this enhancing effect of strength

is not obvious for the 0.3 vol% addition of steel fiber.

Similar test result was also obtained, where the density

and compressive strength of fiber-reinforced concretes

are about the same as those of the plain counterparts

due to 1.0 vol% addition of steel fiber [28].

Due to the expected positive effect of the interfa-

cial bond on the compressive strength [3] for the

same batch of aggregate, different matrix compres-

sive strength should result in different mortar

compressive strength. For fiber reinforced mortars

with low fiber volume content, if the contribution

strength of steel fiber outweighs the decrease of the

matrix compressive strength resulting from the

increase of air content, the compressive strength of

the fiber reinforced mortar may increase due to the

addition of the low content of steel fiber; otherwise,

the addition of the low content of steel fiber may have

little or adverse effect on the compressive strength of

the fiber reinforced mortar. With the increased fiber

volume content, the compressive strength of the fiber

reinforced mortars can greatly increase. For instance,

the 4.67% higher air content of the hardened mortar

and 0.3 vol% steel fiber results in a lower compres-

sive strength of mortar 030003, compared with

mortar 030000 with 4.03% air content; on the other

hand, the 3.47% lower air content of the hardened

mortar and 0.3 vol% steel fiber results in a higher

compressive strength of mortar 051003, compared

with mortar 051000 with 4.84% air content. When the

fiber volume content increases from 0.3 to 1.0 vol%,

such as mortar 030010, although the air content of

hardened mortar increases, the compressive strength

of mortar 030010 increases correspondingly.

3.3 Toughness of the mortar

In the three-point bending strength tests, failure

modes of mortars without and with steel fiber are

Table 4 Compressive and bending strengths of the mortar

Strength at 28 days (MPa) w/b = 0.3 w/b = 0.5

030000 030003 030010 031000 031003 050000 050003 051000 051003

Compressive 72.5 68.9 78.0 78.6 74.9 39.8 34.9 42.6 46.7

Bending 8.9 9.8 11.3 9.0 9.1 7.1 6.8 7.2 7.2

130 Materials and Structures (2010) 43:125–139

quite different, see Fig. 3. Mortar without steel fiber

failed immediately when the maximum elastic load

reached; whereas mortar with steel fiber showed good

ductile behavior. When the maximum elastic load

was reached, the load decreased and the crack width

increased gradually. For the mortar with 1.0 vol%

steel fiber, the load even increased after the reach of

the maximum elastic load. The typical load-flexure

extension (P–d) curves of mortars without and with

steel fiber are schematically shown in Fig. 4.

For each type of the mortar, three companion

specimens were used to determine the bending

strength of the mortar. The maximum elastic load

Pmax-elastic and the corresponding elastic flexure

deflection dmax-elastic of each specimen are summa-

rized in Table 5. Then, the dissipated energy Welastic

and Wpost-elastic before and after Pmax-elastic of each

specimen were calculated, where Welastic and Wpost-

elastic are the area under the load-deflection curve, see

Fig. 4; the total dissipated energy Wtotal is the sum of

Fig. 3 Typical failure

modes of the mortars: awithout steel fiber; b with

steel fiber

Fig. 4 Load deflection

(P–d) curves of specimens:

a without steel fiber; b with

steel fiber

Materials and Structures (2010) 43:125–139 131

the energy Welastic and Wpost-elastic. For the conve-

nience of comparing, the minimum flexure extension

dmin of all specimens with steel fiber was determined

to calculate the dissipated energy Wpost-elastic after

Pmax-elastic, where dmin = 0.3114 mm. The calculated

Welastic, Wpost-elastic and Wtotal of each type of the

mortar can be seen in Table 5.

Then, the influence of factors on the toughness of

the mortar was analyzed, where the toughness is

generally defined as the energy absorption capacity

[28]. It can be seen from Table 5 that the Welastic,

Wpost-elastic and Wtotal of mortars with low water/

binder ratio are larger than those of the counterparts

with larger water/binder ratio, indicating high

toughness of mortars with low water/binder ratio. In

addition, the toughness of the mortar is influenced by

addition of steel fiber. For the mortars with w/b = 0.3,

both the Welastic and Wpost-elastic increase with the

addition of steel fiber. As a result, the Wtotal of mortar

030010 is more than twice of that of mortar 030000

due to the addition of 1.0 vol% steel fiber. For the

mortars with w/b = 0.5, although Welastic is nearly

identical, Wpost-elastic is still increased due to the

addition of steel fiber, resulting in an increase of the

Wtotal. For steel fiber reinforced concrete, the tough-

ness is also remarkably improved by the addition of

1.0 vol% steel fiber [28]. Silica fume, however, has

little or even negative effect on the mortar without

Table 5 Dissipated energy of the mortar

Specimen Dissipated energy (kN�mm) Pmax-elastic (kN) dmax-elastic (mm)

Welastic Wpost-elastic Wtotal

Each Mean value Each Mean value Each Mean value

w/b = 0.3 030000 0.32990 0.3135 0.0391 0.1361 0.3690 0.4496 3.9680 0.1730

0.30070 0.0616 0.3623 3.8676 0.1603

0.31004 0.3076 0.6176 3.8428 0.1614

030003 0.3785 0.4102 0.2926 0.2729 0.6711 0.6831 4.3024 0.1830

0.5442 0.1700 0.7142 4.4272 0.2611

0.3081 0.3561 0.6642 4.0551 0.1554

030010 0.4278 0.4298 0.5416 0.5452 0.9694 0.9750 4.6952 0.1862

0.4125 0.5193 0.9318 4.5721 0.1805

0.4490 0.5748 1.0238 4.6458 0.1946

031000 0.2636 0.2642 0.0653 0.1311 0.3289 0.3953 3.9015 0.1377

0.2700 0.0639 0.3339 3.9160 0.1367

0.2589 0.2641 0.5231 3.9872 0.1296

031003 0.2927 0.2742 0.4915 0.4309 0.7842 0.7051 4.1888 0.1372

0.2524 0.3943 0.6467 3.8903 0.1295

0.2774 0.4068 0.6843 3.9871 0.1340

w/b = 0.5 050000 0.1990 0.2093 0.1342 0.1001 0.3332 0.3094 3.0585 0.1288

0.2375 0.0852 0.3227 3.3676 0.1406

0.1914 0.0809 0.2723 3.0319 0.1287

050003 0.1824 0.2002 0.2272 0.2401 0.4096 0.4402 2.8424 0.1280

0.1746 0.2788 0.4534 2.9023 0.1281

0.2436 0.2141 0.4577 2.9898 0.1673

051000 0.1780 0.1987 0.0436 0.0500 0.2216 0.2486 2.8456 0.1235

0.2160 0.0904 0.3064 3.2396 0.1334

0.2020 0.0159 0.2179 3.1757 0.1292

051003 0.1884 0.2176 0.2377 0.2497 0.4261 0.4673 3.0621 0.1234

0.2377 0.2709 0.5086 3.4266 0.1367

0.2266 0.2405 0.4671 3.0626 0.1472

132 Materials and Structures (2010) 43:125–139

steel fiber. For the steel fiber reinforced mortars, the

Wtotal is increased due to the 10% addition of silica

fume.

3.4 Properties of the interfacial transition zone

in steel fiber reinforced mortars

Nanoindentation tests were carried out to study the

steel fiber–matrix and fiber–matrix-aggregate inter-

faces in mortars 030003, 031003 and 050003. Firstly,

the P–h curves of each indented area were checked to

determine the validity of each indentation. Irregular

P–h curves and curves exhibiting large displacement

jump at any portion of the loading portion were

discarded. The irregular nature of those curves may

be due to the presence of a large void [20] while the

discontinuous load–displacement curves may lie in

the surface cracking during the force-driven indenta-

tion tests [18]. Since the contact stiffness is measured

only at peak load, and no restrictions are placed on

the unloading data being linear during any portion of

the unloading [16], curves showing nonlinear char-

acteristic in the unloading portion were adopted.

After the check of the data validation, the

secondary electron (SE) image of the each indented

area was used to determine the distribution of all

indents in each indented area. The whole indents

were divided into appropriate groups: indents in the

steel fiber, indents in the matrix, and/or indents in the

aggregate. There are also indents in partially hydrated

cement clinkers in some cases. The hardness profiles

of the steel fiber–matrix and steel fiber–matrix-

aggregate interfacial zones of mortars 030003,

031003 and 050003 are shown in Figs. 5, 6, 7, where

all values were determined as a function of the

distance from the fiber surface; the vertical axis

denotes the edge of the actual steel fiber and/or the

aggregate surface is schematically marked.

The hardness profile of the ITZ between steel fiber

and matrix in Fig. 5a shows the trend of type I in

Fig. 2. In the distance 5–35 lm approaching the steel

fiber surface, there is an obvious increase in hardness.

Away the steel fiber surface, i.e. in the 35–65 lm

zones, the hardness becomes relatively constant;

while in Fig. 5b, the bonds across the interfaces are

also efficient, as seen by the obvious rise in the

hardness profile as the steel fiber and aggregate are

approached. Low hardness is observed in the distance

of 20–40 lm away the steel fiber–matrix interface.

Within the steel fiber–matrix-aggregate zone, there

are also some partially hydrated cement clinkers near

the aggregate and fiber surfaces. As a whole, in

mortar 030003, the bonding across the interfaces is

quite efficient.

In Fig. 6a, the whole hardness profile of the ITZ

between steel fiber and matrix likes type III in Fig. 2,

but it is more uniform. Due to the addition of the

silica fume, the hardness profile of the steel fiber–

matrix interfacial zone is quite different from that of

the mortar 030003. It is interesting to note that there

is no obvious trough in the interfacial zone in Fig. 6a;

comparatively low hardness is observed just in the

10–30 lm zones. Compared with 0.1636 GPa at the

weakest point in Fig. 5a, this value reaches

0.2391 GPa in Fig. 6a, indicating an increase of the

hardness value at the weakest point owing to the 10%

addition of the silica fume. In Fig. 6b, small rise in

the hardness profile as the steel fiber is approached

indicates good bonding across the steel fiber-matrix

interface. However, this bond is not as efficient as

that of the steel fiber-matrix interface shown in

Fig. 5b; the hardness value at the weakest point in

Fig. 6b is very low, just 0.034 GPa. The influence of

the silica fume is not as obvious as that shown in

Fig. 6a. With the approaching of the aggregate, rise

in the profiles is observed, indicating at least partial

bonding at the aggregate–matrix interface, as sug-

gested by Igarashi et al. [13]. As a whole, the bond

behavior in 031003 is good, but not as efficient as that

in 030003. The addition of silica fume results in no

distinct presence of weak ITZ between steel fiber and

matrix in some parts of the mortar; however, in the

steel fiber–matrix-aggregate interfacial zone, the

effect of the silica fume is not obvious.

The hardness profile of the ITZ in Fig. 7a shows

typical characteristic of type IV in Fig. 2. Compared

with the comparatively high matrix hardness in

mortars 030003 and 031003, comparatively low

matrix hardness in mortar 050003 is observed due

to the larger water/binder ratio. The lowest hardness

is observed near the steel fiber surface, just 25.6 lm

from the fiber surface. This value is 0.1452 GPa,

indicating a decrease of the hardness at the weakest

point owing to the increase of the water/binder ratio.

A trough is shown in the 0–35 lm zones. This

indicates that the interfacial bonding at the actual

interface in specimen 050003 is poor and not as

effective as that obtained in specimens 030003 and

Materials and Structures (2010) 43:125–139 133

031003 with low water/binder ratio. In Fig. 7b, small

rise in the hardness profile as the steel fiber surface is

approached can be indicative of at least partial

bonding at the interface. Unlike the cases in mortars

030003 and 031003, the smallest hardness value

occurred near the fiber surface. It is very surprising to

note that in the distance 5–20 lm from the aggregate

surface, the hardness are quite similar to those of the

aggregate although it is part of matrix. It is doubted

that there maybe a piece of aggregate underneath this

area. As a whole, partial bond exhibits at the

interfaces in mortar 050003.

3.5 Relationship between the ITZ

and macroscopic mechanical properties

As pointed out above, the addition of the 1.0 vol%

ductile steel fiber can greatly improve the three-point

bending strength and fracture toughness of cementi-

tious materials even the compressive strength is little

0

1

2

3

4

5

6

7

8

9

10

11

-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70

Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Steel fiber

Matrix

(a)

01234567

89

101112131415

-30 -20 -10 0 10 20 30 40 50 60 70 80 90

Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Steel fiberMatrix

Partially hydrated cementAggregate

(b)

Aggregate surface

Fig. 5 Hardness profile in

the interfacial transition

zones of mortar 030003:

a steel fiber–matrix

interfacial zone; b steel

fiber–matrix-aggregate

interfacial zone

134 Materials and Structures (2010) 43:125–139

improved [28]. The magnitude of the bending

strength and fracture toughness increase of fiber-

reinforced cementitious composites is mainly con-

trolled by their interfacial characteristics and

microstructural morphology near the fiber surface.

Thus, the effect of interfacial characteristics and

microstructural morphology near the fiber surface on

the bending strength and toughness of the fiber

reinforced mortar are mainly discussed.

For mortar without and with silica fume, such as

030003 and 031003, different bond conditions in the

ITZ lead to different bending strength and fracture

toughness. Although efficient bonding across the steel

fiber–matrix interface is shown in both 030003 and

031003, in the steel fiber–matrix-aggregate zone, the

situation is different. In Fig. 5, efficient bonding

across the interfaces in 030003 is observed. Due to

the largest air content in mortar 031003, the thickness

of the ITZ in the steel fiber–matrix-aggregate zone is

very wide, approximate 60 lm; also, there are many

voids and microcracks in this zone, see the BSE

image in Fig. 8. Some microcracks even extend from

the fiber surface to the aggregate surface. The

comparative large and discontinuous pores may be

0

1

2

3

4

5

6

7

8

9

10

11

-30 -20 -10 0 10 20 30 40 50 60 70 80

Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Matrix

Steel fiber

(a)

0123456789

101112131415

-30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Steel fiberMatrixAggregate

(b)

Aggregate surface

Fig. 6 Hardness profile in

the interfacial transition

zones of mortar 031003:

a steel fiber–matrix

interfacial zone; b steel

fiber–matrix-aggregate

interfacial zone

Materials and Structures (2010) 43:125–139 135

related to the consumption of calcium hydroxide

crystals by reaction with silica fume while the

increase in microcracks may result from autogenous

shrinkage in the presence of comparatively large

amounts of silica fume [29]. Those porosity and

microcracks may result in earlier initiation and

development of cracks in this zone. As a result,

compared with mortar 030003, comparatively low

bending strength and Welastic are obtained in mortar

031003.

For mortar with different water/binder ratio, such

as 030003 and 050003, bending strength and fracture

toughness differ a lot due to the different bond

characteristic in the interfaces. In mortar 030003, the

rise in the hardness profiles as the steel fiber and

aggregate are approached indicates efficient bonds

across the steel fiber–matrix and fiber–matrix-aggre-

gate interfaces; while in mortar 050003, due to the

large water/binder ratio, a trough is shown in the

distance 0–35 lm from the steel fiber surface in the

steel fiber–matrix interfacial zone or in the distance

5–40 lm from the steel fiber surface in the fiber–

matrix-aggregate interfacial zone, indicating poor

bond in those zones. From the BSE image shown in

0

1

2

3

4

5

6

7

8

9

10

11

-20 -10 0 10 20 30 40 50 60 70Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Matrix

Steel fiber

(a)

(b)

0123

456789

1011

12131415

-30 -20 -10 0 10 20 30 40 50 60 70 80 90 100

Distance from fiber surface (µm)

H-h

ardn

ess

(GPa

)

Steel fiberMatrixAggregate

Aggregate surface

Fig. 7 Hardness profile in

the interfacial transition

zones of mortar 050003:

a steel fiber–matrix

interfacial zone; b steel

fiber–matrix-aggregate

interfacial zone

136 Materials and Structures (2010) 43:125–139

Fig. 9, a clear, porous transition zone and discon-

tinuous bleeding voids underneath the fiber are

observed. Due to those reasons, the bending strength

and fracture toughness of mortar 050003 are lower

than those of mortar 030003.

4 Conclusions

The following concluding remarks can be made from

this experimental work:

(1) Properties of the fresh mortar are greatly

influence by water/binder ratio, the addition of

silica fume and steel fiber. With the increase of

the water/binder ratio, the yield stress and

plastic viscosity decrease corresponding. Due

to the addition of the silica fume, the yield stress

increases more for mortars with low water/

binder ratio. For mortars with low water/binder

ratio, yield stress and plastic viscosity increase

greatly resulting from the addition of steel fiber;

while for mortars with large water/binder ratio,

the workability and consistency is less influ-

enced by the low volume addition of steel fiber.

(2) At both water/binder ratios, compressive strength

increased with silica fume. If the air content of the

hardened mortar is well controlled, compressive

strength can be increased due to the low volume

(0.3 vol%) addition of steel fiber; otherwise, the

addition of the low steel fiber content may have

little or adverse effect on the increase of the

compressive strength of the mortar. With the

increased volume addition of steel fiber from 0.3

to 1.0 vol%, the compressive strength of the

mortar is remarkably increased.

(3) Failure modes of the mortar are influence by the

addition of steel fiber. Good ductile behavior is

shown by steel fiber reinforced mortars. The

toughness of the mortar is denoted by the

dissipated energy. Mortars with lower water/

binder ratio show higher toughness. For mortars

with low water/binder ratio, the toughness is

increased with the increased volume addition of

steel fiber; for mortars with large water/binder

ratio, the elastic energy is nearly identical while

the post elastic energy is still increased due to

the addition of steel fiber.

(4) The three-point bending strength and the tough-

ness of the steel fiber reinforced mortar are

related to the interfacial characteristics and

microstructural morphology near the fiber sur-

face. For mortar with the same water/binder

ratio, comparatively low bending strength and

Welastic in mortar with 10% silica fume result

from the wider thickness of the ITZ in the steel

fiber–matrix-aggregate interfacial zone and

voids and microcracks within this zone. Those

discontinuous pores and the increase in micro-

cracks may result from the reaction with silica

fume and autogenous shrinkage. For mortar

with different water/binder ratio, due to the poor

bonds across the steel fiber–matrix and fiber–

matrix-aggregate interfaces in mortar with

higher water/binder ratio, the bending strength

Fig. 8 Voids and microcracks in the steel fiber–matrix-

aggregate zone in mortar 031003

Fig. 9 Bleeding void and porous zone near the surface of steel

fiber in mortar 050003

Materials and Structures (2010) 43:125–139 137

and fracture toughness are obviously lower than

those of mortar with lower water/binder ratio.

The present work aims at linking the nano-

mechanical properties of the ITZ with the macro-

scopic mechanical properties. It will help to proceed

in the development of more ductile cement based

materials and obtain fiber reinforced cementitious

composites with high tensile strength, flexural

strength and fracture toughness by tailoring the

microstructure of the fiber–matrix interface and

fiber–matrix-aggregate interface.

Acknowledgements This research was done at the

Norwegian University of Science and Technology (NTNU),

when the first author is working at the Department of Structural

Engineering as a visiting Professor for one year. The first

author gratefully thank the support by the Norwegian Research

Councils a part of the Cultural Agreement between Norway

and China—Government scholarships 2007/2008 (No.

26X35003) and the National Natural Science Foundation of

China (No. 50508020). The invitation provided by Professor

Stefan Jacobsen in NTNU to cooperate with this research work

is gratefully appreciated. In addition, the financial support from

the COIN (Concrete Innovation Centre) Project 3-3.5—

Nanotechnology applied to Cement-based Materials is greatly

appreciated. Help given by Dr. Hilde Lea Lein, engineer Ove

Edvard Loraas, engineer Arild Monsøy and Wilhelm Dall in

NTNU is also acknowledged.

References

1. Mitsui K, Li Z, Lange DA et al (1994) Relationship

between microstructure and mechanical properties of the

paste-aggregate interface. ACI Mater J 91(1):30–39

2. Husem M (2003) The effects of bond strengths between

lightweight and ordinary aggregate-mortar, aggregate-

cement paste on the mechanical properties of concrete.

Mater Sci Eng A 363(1–2):152–158. doi:10.1016/S0921-

5093(03)00595-1

3. Akcaoglu T, Tokyay M, Celik T (2005) Assessing the ITZ

microcracking via scanning electron microscope and its

effect on the failure behavior of concrete. Cement Concr

Res 35(2):358–363. doi:10.1016/j.cemconres.2004.05.042

4. Bentur A, Alexander MG, Bentz D et al (2000) Review of

the work of the RILEM TC 159-ETC: engineering of the

interfacial transition zone in cementitious composites.

Mater Struct 33(2):82–87. doi:10.1007/BF02484160

5. Merchant IJ, Macphee DE, Chandler HW et al (2001)

Toughening cement-based materials through the control of

interfacial bonding. Cement Concr Res 31(12):1873–1880.

doi:10.1016/S0008-8846(01)00500-2

6. Guinea GV, El-Sayed K, Rocco CG et al (2002) The effect

of the bond between the matrix and the aggregates on the

cracking mechanism and fracture parameters of concrete.

Cement Concr Res 32(12):1961–1970. doi:10.1016/S0008-

8846(02)00902-X

7. Prokopski G, Langier B (2000) Effect of water/cement

ratio and silica fume addition on the fracture toughness and

morphology of fractured surfaces of gravel concretes.

Cement Concr Res 30(9):1427–1433. doi:10.1016/S0008-

8846(00)00332-X

8. Mouret M, Bascoul A, Escadeillas G (1999) Microstruc-

tural features of concrete in relation to initial temperature—

SEM and ESEM characterization. Cement Concr Res

29(3):369–375. doi:10.1016/S0008-8846(98)00160-4

9. Gatty L, Bonnamy S, Feylessoufi A et al (2001) A trans-

mission electron microscopy study of interfaces and matrix

homogeneity in ultra-high-performance cement-based

materials. J Mater Sci 36(16):4013–4026. doi:10.1023/A:

1017938725748

10. Diamond S (2001) Considerations in image analysis as

applied to investigations of the ITZ in concrete. Cement

Concr Compos 23(2–3):171–178. doi:10.1016/S0958-9465

(00)00085-8

11. Scrivener KL (2004) Backscattered electron imaging of

cementitious microstructures: understanding and quantifi-

cation. Cement Concr Compos 26(8):935–945. doi:

10.1016/j.cemconcomp.2004.02.029

12. Sun W, Mandel JA, Said S (1986) Study of the interface

strength in steel fiber-reinforced cement-based composites.

J Am Concr Inst 83(4):597–605

13. Igarashi S, Bentur A, Mindess S (1996) Microhardness

testing of cementitious materials. Adv Cem Based Mater

4(2):48–57

14. Cross WM, Sabnis KH, Kjerengtroen L et al (2000)

Microhardness testing of fiber-reinforced cement paste.

ACI Mater J 97(2):162–167

15. Velez K, Sorrentino F (2001) Characterization of cemen-

titious materials by nanoindentation. In: Kurdowski W,

Gawlicki M (eds) Kurdowski symposium—science of

cement and concrete, Krakow, June 20–21, pp 67–77

16. Oliver WC, Pharr GM (1992) Improved technique for

determining hardness and elastic modulus using load and

displacement sensing indentation experiments. J Mater Res

7(6):1564–1580. doi:10.1557/JMR.1992.1564

17. Constantinides G, Ulm FJ, Van Vliet K (2003) On the use

of nanoindentation for cementitious materials. Mater Struct

36(3):191–196. doi:10.1007/BF02479557

18. DeJong MJ, Ulm FJ (2007) The nanogranular behavior of

C-S-H at elevated temperatures (up to 700�C). Cement

Concr Res 37(1):1–12. doi:10.1016/j.cemconres.2006.09.

006

19. Mondal P, Shah SP, Marks L (2007) A reliable technique

to determine the local mechanical properties at the nano-

scale for cementitious materials. Cement Concr Res

37(10):1440–1444. doi:10.1016/j.cemconres.2007.07.001

20. Mondal P, Shah SP, Marks LD (2008) Nanoscale charac-

terization of cementitious materials. ACI Mater J 105(2):

174–179

21. Nemecek J, Kabele P, Bittnar Z (2004) Nanoindentation

based assessment of micromechanical properties of fiber

reinforced cementitious composite. In: 6th RILEM sym-

posium on fiber-reinforced concrete (FRC), BEFIB,

Varenna, Italy, pp 401–410

22. Zhu W, Sonebi M, Bartos PJM (2004) Bond and interfacial

properties of reinforcement in self-compacting concrete.

Mater Struct 37(7):442–448. doi:10.1007/BF02481580

138 Materials and Structures (2010) 43:125–139

23. Banfill PFG (1994) Rheological methods for assessing the

flow properties of mortar and related materials. Constr

Build Mater 8(1):43–50. doi:10.1016/0950-0618(94)9000

7-8

24. Faroug F, Szwabowski J, Wild S (1999) Influence of su-

perplasticizers on workability of concrete. J Mater Civ Eng

11(2):151–157. doi:10.1061/(ASCE)0899-1561(1999)11:2

(151)

25. Jacoben S, Arntsen B (2008) Aggregate packing and -void

saturation in mortar and concrete proportioning. Mater

Struct 41(4):703–716. doi:10.1617/s11527-007-9275-4

26. Kjellsen KO, Wallevik OH, Hallgren M (1999) On the

compressive strength development of high-performance

concrete and paste—effect of silica fume. Mater Struct

32(1):63–69. doi:10.1007/BF02480414

27. Bentur A (2000) Role of interfaces in controlling durability

of fiber-reinforced cements. J Mater Civ Eng 12(1):2–7.

doi:10.1061/(ASCE)0899-1561(2000)12:1(2)

28. Balendran RV, Zhou FP, Nadeem A et al (2002) Influence

of steel fibres on strength and ductility of normal and

lightweight high strength concrete. Build Environ

37(12):1361–1367. doi:10.1016/S0360-1323(01)00109-3

29. Feldman RF, Huang C (1985) Properties of Portland

cement-silicate fume pastes-I porosity and surface prop-

erties. Cement Concr Res 15(5):765–774. doi:10.1016/

0008-8846(85)90141-3

Materials and Structures (2010) 43:125–139 139