- Email: [email protected]
Nanoindentation study of the interfacial zone between cellulose fiber and cement matrix in extruded composites
Accepted Manuscript Nanoindentation study of the interfacial zone between cellulose fiber and cement matrix in extruded composites R.S. Teixeira, G.H.D. Tonoli, S.F. Santos, E. Rayon, V. Amigó, H. Savastano, Jr., F.A. Rocco Lahr PII:
S0958-9465(16)30464-4
DOI:
10.1016/j.cemconcomp.2017.09.018
Reference:
CECO 2916
To appear in:
Cement and Concrete Composites
Received Date: 19 August 2016 Revised Date:
14 July 2017
Accepted Date: 29 September 2017
Please cite this article as: R.S. Teixeira, G.H.D. Tonoli, S.F. Santos, E. Rayon, V. Amigó, H. Savastano Jr., F.A.R. Lahr, Nanoindentation study of the interfacial zone between cellulose fiber and cement matrix in extruded composites, Cement and Concrete Composites (2017), doi: 10.1016/ j.cemconcomp.2017.09.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Nanoindentation study of the interfacial zone between cellulose fiber and
2
cement matrix in extruded composites
3 4
R.S. Teixeira1; G.H.D. Tonoli2; S.F. Santos3, E. Rayon4, V. Amigó4; H. Savastano Jr5; F.A.
5
Rocco Lahr1. 1
University of São Paulo (USP), Department of Structural Engineering. School of Engineering of São Carlos, Brazil; 2 Federal University of Lavras (UFLA), Department of Forest Science, Brazil; 3 São Paulo State Universit (UNESP), School of Engineering, Department of Materials and Technology, Brazil; 4 Universitat Politècnica de València (UPV), Department of Mechanical and Materials Engineering, Spain. 5 University of São Paulo (USP), Department of Biosystems Engineering, Brazil.
15
ABSTRACT
16
The present study shows the application of the nanoindentation technique to evaluate the
17
properties of the cellulose fiber-cement matrix interfacial zone in composites prepared with an
18
auger extruder. The degree of strength of the bond between fiber and matrix is recognized as
19
important variable that influences macro-mechanical properties, such as modulus of rupture
20
and toughness of cement based composites. The nanoindentation measurements showed the
21
highest hardness and elastic modulus in the part inner of the cellulosic fiber after hydration
22
process due to precipitation and re-precipitation of cement hydration products. These results
23
indicate that mineralization of the cellulosic fibers can affect the stress distribution and
24
interfacial bond strength in the cement based composite.
TE D
M AN U
SC
RI PT
6 7 8 9 10 11 12 13 14
EP
25
KEYWORDS: vegetable fiber, microstructure, mechanical properties, ordinary Portland
27
cement, transition zone.
28
AC C
26
29
1 INTRODUCTION
30
It is widely accepted that concrete at the mesoscopic scale should be treated as a three-phase
31
composite material constituted by bulk hardened cement paste, aggregates and an interfacial
32
transition zone (ITZ) between these two. [1] Therefore, traditional ITZ is the region of the
33
cement matrix around the aggregate particles, which is perturbed by the presence of the
34
aggregate. Its origin lies in the packing of the cement grains against the much larger
35
aggregate, which leads to a local increase in porosity and predominance of smaller cement
36
particles in this region. The ITZ is region of gradual transition and is highly heterogeneous,
37
nevertheless the average microstructural features may be measured [2].
ACCEPTED MANUSCRIPT
The cementitious composites reinforced with cellulosic fibers are complex materials. Thus,
39
cement based composite design must be considered for the interrelation between interfacial
40
transition zone, fibers, matrix and processing method. Particle and fiber size distribution,
41
particles and fibers packing and rheological behavior of the cement paste with fibers in the
42
fresh state also interfere on the interfacial transition zone of the hardened cement based
43
composite. Furthermore, changes in the mechanical properties over time can occur due to
44
microstructural changes in the fiber-matrix interface and bulk, as a consequence of the
45
continued hydration process that occurs in the fiber surroundings, i.e. if the interfacial
46
transition zone becomes denser due to re-precipitation of cement hydration products in the
47
porous layers. [3]. Therefore, in this manuscript an interfacial transition zone (ITZ)
48
terminology is related to a region between bulk hardened cement paste and cellulosic fiber.
49
Kraft pulp fiber-reinforced cement-based materials have being increasingly used in most
50
developing countries as construction materials [4–6]. Fibers usually increase mechanical
51
strength and the capacity of cement based composite to absorb energy through the distribution
52
of microcracks into the material. The bonding between cellulose fiber and Portland cement
53
plays a major role in the transference of the load from the matrix to the reinforcement,
54
impacting the mechanical behavior of the resulting fiber-cement material. This bonding can
55
be of physical (mechanical interlocking) or chemical (mainly hydrogen bonds) nature, or a
56
combination of both [7].
57
Extrusion process has a potential of producing fiber reinforced cementitious composites with
58
higher performance compared to the traditional slurry dewatering processes such as Hatschek
59
and Flow-on methods. Several advantages of extrusion moulding can be pointed for
60
manufacturing of fibrous composites: (i) low implementation cost; (ii) low capillary porosity
61
of extruded composites; and (iii) low water/cement ratio, which plays a significant role for
62
improving of the fiber-matrix interface zone [8–10]. As mentioned above, the control of fiber-
63
matrix interface zone in extruded composites depends on cellulosic fibers and matrix
64
constituent’s properties, cure procedure as well as weathering conditions that the composite
65
are exposed. Previous studies highlighted that water/cement ratio influences the interface zone
66
thickness and that lower porosity may increase composite strength and matrix toughness [9–
67
12]. Therefore, the optimization of the interfacial zone is not an easy task since it is necessary
68
to satisfy the equilibrium between fiber-matrix adhesion and fiber pull-out. In addition, there
69
is still a lack of information about the consequences of the different types of interface zone
70
between cellulose fibers and matrix, on composite mechanical properties. Thus, an effort is
71
required in order to understand the mechanical properties at fiber-matrix transition zone.
AC C
EP
TE D
M AN U
SC
RI PT
38
ACCEPTED MANUSCRIPT
The macroscale mechanical properties of bulk materials are dependent on the properties of
73
their microscale/nanoscale constituents and the interface between these constituents.
74
Microindentation technique is one of the developing tests that have been used for evaluating
75
mechanical characteristics of microsized zones in diverse cement based composites [13–15].
76
Nevertheless, microindentation is not capable to detect mechanical differences at the
77
interfaces between cement matrix and reinforced fibers, due to some limitations, such as a
78
resolution of microindentation system in terms of load is about 5.0 mN, to analyze some
79
structural features because of the heterogenity of the specimen, inadequate spacing between
80
indents or form edges, operator’s visual acuity. Quasistatic nanoindentation has become the
81
standard technique used for nanomechanical characterization of small volumes of material,
82
such as elastic modulus, hardness at the nanoscale. It is used machines that can record small
83
load and displacement with high accuracy and precision, during loading and unloading can be
84
continuously monitored with high resolution in the nanonewton and nanometer ranges. In this
85
technique is used usually the Berkovich tip is a 3-sided pyramid that have an apex appropriate
86
for investigating the interface transition zone between fiber and matrix. Therefore,
87
nanoindentation technique can contribute to quantitatively measure the underlying source of
88
the bulk response at the nanoscale in cement based composites that are heterogeneous
89
material from both a chemical composition and nanoscale structure standpoint. Being able to
90
evaluate at this scale allows micro and nano structural "tuning" to be performed to optimize
91
bulk properties [16,17]. The aim of this study was to evaluate the interfacial zone between
92
fibers to cement matrix in extruded composites reinforced with eucalyptus pulp, using a
93
nanoindentation technique allied with microscopic examinations.
94
EP
TE D
M AN U
SC
RI PT
72
2 EXPERIMENTAL
96
2.1 Materials
97
Unbleached unrefined eucalyptus (Eucalyptus grandis) Kraft pulp was provided by Fibria
98
S/A, Brazil. The cellulose pulp was collected directly from the mill, prior to drying and
99
pressing. It was extensively washed with water and centrifuged to remove any residual
AC C
95
100
chemicals from the pulping processes.
101
Eucalyptus pulps were analyzed about morphology (length and width) and concentration of
102
particles (fibers per gram and content of fines). A fine element (small particles with length
103
lower than 200 µm that are commonly called fines, which also contain band-like materials
104
from both primary wall and secondary wall layers of the fibers) was detected in the pulp with
105
dimensions lower than those of fibers. Fines do not contribute significantly to cement based
ACCEPTED MANUSCRIPT
106
composite strength but act more as filler [18]. Physical characteristics of the unbleached
107
eucalyptus pulp are showed in Table 1.
108 2.2 Composite mix design
110
Formulation used in the composite production was based on previous study [8], and is
111
described in Table 2.
112
Portland cement and limestone particles distribution was performed by laser particle size
113
analyzer (Malvern Mastersizer S long bed, version 2.19). Cement and limestone particles
114
showed 50% of their size below 11 µm and 16 µm respectively.
115
Hydroxypropylmethylcellulose (HPMC) from Aditex, is a superplasticizer which disperses
116
cement particles, reducing the flow stress and the viscosity of the cement paste. Polyether
117
carboxylic provided by Grace, commercially called ADVA 170, is an additive which reduces
118
the surface tension of the water and thus causes an increase in the surface area of the liquid
119
which improves the water absorption, forming a lubrication layer between the wall of the
120
extruder and the surface of the fiber cement board. HPMC with 86,000 average molecular
121
weight and 5.39 cPs viscosity (at 2% concentration in water at 20°C) and polyether carboxylic
122
commercially named ADVA 170, were used as rheological modifiers in fiber-cement, with
123
dosage of 1 wt% of the total mass of the particulate inorganic materials (Portland cement and
124
limestone) to promote the pseudo-plastic behavior of the composite, which, in turn, enabled
125
the extrusion process.
126
TE D
M AN U
SC
RI PT
109
2.3 Composite extrusion
128
The mass was mixed in a mechanical Eirich intensive mixer (capacity of 10 L) for 5 min at
129
125 rpm, 5 min at medium velocity (220 rpm) and finally 5 min at high velocity (450 rpm).
130
The water-cement (w/c) ratio was 0.33. The composite mixture was transferred to a Verdes
131
051 extruder (Figure 1a-b). The speed of the extruder was 4 mm/s; die width/diameter ratio
132
was 3.3. Composites were molded in 15 mm x 50 mm x 200 mm plates. The illustration of the
133
chamber parts of the extruder is presented in Figure 1c.
134
The composite mixture was re-circulated into the extruder for 5 min before tailoring the
135
samples. Pads with 15 mm x 50 mm x 200 mm were cured at water vapor saturated
136
environment (in sealed plastic bags) at 25 ± 2ºC for two days. Subsequently, the specimens
137
were maintained in water vapor saturated environment (in sealed plastic bags) and placed in a
138
chamber at 45ºC for five days (thermal curing) totalizing 7 days of cure [21].
139 140
AC C
EP
127
ACCEPTED MANUSCRIPT
2.4 Microstructure characterization
142
The specimens were cut out from the middle section of a larger volume of fiber cement, with
143
minimal deformation, by using a precision cutter, Buehler, model Isomet, with precision
144
diamond saw, 15LC, and using an isopropyl alcohol based cutting fluid to cut. This method
145
provides roughness smaller. The specimens were embedded in epoxy resin (MC-
146
DUR1264FF) previously to be visualized in a scanning electron microscope (SEM). A final
147
polishing was carried out using sequentially 8-4 µm, 4-2 µm and 1-0 µm diamond abrasives,
148
as reported elsewhere [22,23]. The polished samples were carbon coated before being
149
analyzed in a JEOL JSM6300 electron microscope with acceleration voltage of 20 kV. A
150
back-scattered electron (BSE) detector was used to observe the morphological features on the
151
cross-section and polished surfaces. BSE imaging permits the identification of cementitious
152
phases since the electron scattering change. Dark and light areas are related to lighter and
153
heavier elements respectively. Energy dispersive spectrometry (EDS) analysis was performed
154
in order to identify the chemical composition in different spots on the same polished surface
155
specimens.
M AN U
SC
RI PT
141
156
2.4.1 Atomic force microscopy characterization (AFM)
158
In contrast to ductile material, cement based composite have regions with higher hardness,
159
lower ductility, and a basically brittle nature. Due to the brittleness, cellulosic fibers and
160
porosity of the composite it is quite often difficult to cut and polish. All areas of the
161
specimens observed by nanoindentation were cut and polished with uniformity and carefully
162
to avoid any pullout of fibers, but as the composite is heterogeneous the response of each
163
region was different according to its resistance. Specimens for AFM characterization were cut
164
with same system mentioned above and embedded in polyacrylamide. The semi-automatic
165
polisher was used to adjust polishing velocity and pressure, i.e. all regions were submitted to
166
same conditions of preparation. Then, the dried surfaces were grinded in a sequence of SiC
167
papers (grit #500, #1000, #2000 and #4000) and polished in a cloth sequentially with 9 µm,
168
6 µm and 3 µm diamond sprays. Finally, the specimen surfaces were cleaned by washing in
169
alcohol and ultrasonic bathing.
170
The characterization of the fiber-matrix interface zone was carried out using the tapping-mode
171
atomic force microscopy (TM-AFM) to obtain height and phase imaging data simultaneously,
172
with Nanoscope III-Digital Instruments. Silicon cantilever with spring constant of around
173
70 N/m and a scan area of 15 µm x 15 µm was used. Samples were prepared in the laboratory
174
environment with temperature around 25°C and relative humidity between 50% and 65%.
AC C
EP
TE D
157
ACCEPTED MANUSCRIPT
175
AFM images were obtained in the same laboratory environment conditions. The test locations
176
were guided by means of an incorporated optical light microscope.
177 2.5 Nanoindentation tests
179
Specimens for nanoindentation tests were prepared the same methodology that specimens to
180
AFM characterizations. Nanoindentation was performed in an Agilent G-200 instrument
181
working in continuous stiffness measurement (CSM), configured at 2 nm of harmonic
182
amplitude, 45 Hz frequency, and using a Berkovich indenter. The test locations were guided
183
by means of an incorporated optical light microscope. To rule out the influence of surface
184
morphology, the indentation depth should be much greater than the characteristic size of
185
surface roughness to avoid errors and size effect in the calculations [17]. Tests were
186
programmed to reach from 100 to 300 nm-indented depths until the maximum depth of 1000
187
nm. Therefore, the depth of the imprint spots used in the nanoindentation was to minimize
188
surface roughness of the samples. Analyses were performed with 8 nm distance between the
189
measurement points [17]. It was performed more than 20 analyses for each region of the
190
composite (bulk matrix, interface or transition zone, anhydrous grain and cellulose fibers).
191
To guarantee a significance results for nanoindentation tests were choose under a completely
192
randomized design the ITZ areas to analyze a linear sequence with four regions (transition
193
zone, cellulose fiber, transition zone, bulk matrix and anhydrous grain or follow sequence:
194
anhydrous grain, bulk matrix, cellulose fiber, transition zone) and twenty replications. The
195
effects of hardness and elastic modulus were compared via statistical contrast by student test
196
at 5% of significance [17]. The ACTION statistical program was used. Therefore, the
197
interfaces towards fibers and matrix were studied in greater detail.
SC
M AN U
TE D
EP
198
RI PT
178
3 RESULTS AND DISCUSSION
200
3.1 Chemical and microstructure characterization
201
Figure 2 shows SEM-BSE micrographs of the extruded composites. Energy dispersive X-ray
202
spectroscopy (EDS) analysis (Figure 2a) indicate that fiber-matrix interfacial zone presents
203
high Ca/Si ratio (between 4 to 5), indicating the possible correspondence with Ca(OH)2
204
crystals with carbonated products, such as calcite, as corroborated by AFM analysis (Figure
205
3) and according to Ruiz-Agudo et al. [24]. This interfacial zone rich in Ca(OH)2 crystals has
206
been widely reported in literature for cementitious composites [25–29]. Bulk matrix revealed
207
an expected lower Ca/Si ratio (approximately 2.5) as observed in Figure 2b. The phase
208
composition of the anhydrous grains and matrix are shown in Figures 4a and 4b, respectively.
209
Spot B presents higher concentration of SiO2 and CaO due to the hydration products of
AC C
199
210
ACCEPTED MANUSCRIPT
cement in relation to the spot A (region of the anhydrous grain).
211 3.2 Mechanical characterization by nanoindentation
213
Figures 5a and 6a present the micrographs of the nanoindentation imprints done along the
214
bulk section of the composites. The spots numbered between 1-10 (Figure 5a), 25-61 (Figure
215
5a) and 81-90 (Figure 6a) represent the bulk matrix of the composite. The spots 15-16 (Figure
216
5a), 20-22 (Figure 5a) and 101-110 (Figure 6a) represents the interfacial zone. The spots 16-
217
19 (Figure 5a) and 91-100 (Figure 6a) represents the cellulosic fiber, while 62-70 (Figure 5a)
218
and 81-90 (Figure 6a) are spots related to anhydrous grains. The individual values of elastic
219
modulus (E) acquired for each spot are summarized in Figures 5b and 6b; while the individual
220
values of hardness (H) is presented in Figures 5c and 6c.
221
Table 3 shows the average and standard deviation values of elastic modulus and hardness
222
obtained from nanoindentation in the different regions studied. The indent spots in the bulk
223
matrix led to elastic modulus of around 70 GPa and hardness of around 5 GPa. These results
224
are similar to those found by Silva et al. [17], who studied nanoindentation in concrete
225
specimens. The bulk matrix of composites based on Portland cement normally presents
226
calcium silicate hydrates (C-S-H) and calcite phases as main components, which have higher
227
rigidity and relative lower porosity than other phases such as Ca(OH)2 and ettringite that are
228
formed by poorly packed structures. This is why the spots correspondent to the
229
interfacial/transition zone presents lower elastic modulus (~17 GPa) and hardness (~0.4 GPa).
230
These results of the elastic modulus and hardness vary in relation to the distance of the
231
transition zone, as showed in Figure 5. Similar result was found by Wang et al. [30] that
232
studied nanoindentation technique in concrete reinforced with steel fibers. The values of
233
hardness in the 11-12, 23-26 indent spots near to the interfacial zone are smaller than those in
234
the matrix region, which are also related to the higher porosity in the interfacial zone,
235
resultant of the low adhesion between fiber and matrix [31,32].
236
Other important observation from this study in nanoscale level is related to the hydration of
237
the cement grains in the composite. The incorporation of water during the mixing of the
238
composite constituents leads to cement hydrate grains that have much lower average atomic
239
numbers than cement grains that were not completely hydrated (anhydrous grains). Then,
240
strong contrast is obtained between unreacted (anhydrous) and reacted material (hydrates)
241
[33]. Anhydrous grain shown in Table 3 presents elastic modulus of around 90.0 GPa and
242
hardness of around 4.5 GPa, which was not significantly different from average values of
243
hydrated bulk matrix. Velez et al. [34] also studied nanoindentation in cement and did not
AC C
EP
TE D
M AN U
SC
RI PT
212
ACCEPTED MANUSCRIPT
find differences between hardness and elastic modulus in anhydrous phases such as alite
245
(C3S), calcium aluminate (C3A) and belite (C2S).
246
The highly alkaline pore water within the fiber-cement interface can induce the cellulose
247
fibers stiffening. It occurs because cellulose fibers are hydrophilic and absorb water rich in
248
cement hydration products, leading to the mineralization phenomenon proposed elsewhere
249
[35–37]. This phenomenon is related to hydrophilic nature of the cellulose fibers. Therefore,
250
this phenomenon is related to the hydrophilic nature of the cellulose fibers. This nature
251
depends on surface nature, surface porosity and lumens the inner of the cellulose fiber. The
252
internal humidity in the fiber can promote a continued hydration process, with precipitation
253
and re-precipitation internally, such as C-S-H, portlandite and calcites, as suggest by Figures
254
7a and 7b. Figure 7a shows hydration products within the cell wall of cellulosic fibers, while
255
the EDS spectra (Figure 7b) in the spots 1 (upper) and 2 (bottom) proves the large amounts of
256
Si and Ca into the cellulose fibers. It explains the higher values of elastic modulus (~144 GPa)
257
and hardness (~9 GPa) observed for the nanoindentation spots related to the cellulosic fiber
258
regions. Therefore, the average elastic modulus of mineralized cellulose fibers can reach
259
higher values than cement bulk matrix. The increase of stiffness of the cellulose fibers is a
260
drawback because can induce the loss of their reinforcement capacity. The efficiency of stress
261
transfer from matrix to fiber that is of fundamental importance in determining the mechanical
262
properties of short fibre reinforced composite is prejudiced. This result explain the premature
263
loss of ductility/toughness and increase of stiffness of composites reinforced with cellulose
264
fibers widely reported in literature [22,23,38,39].
265
Therefore, the fiber to matrix bonding is impaired because cellulose fiber tends to exchange
266
water with the matrix system by capillary action, leading to high dimensional variation and
267
detachment of the fiber surface from the matrix, as showed in Figure 7a, creating voids that
268
explain the low results of elastic modulus and hardness in the transition zone. Finally, the
269
present observations may state that an optimal situation would be the cellulose fibers
270
protection from excessive water absorption, what can be achieved by less hydrophilic surfaces
271
(either natural or after an adequate treatment) and without losing the fiber bridging. Fiber
272
bridging is important to guarantee the expend of energy during fiber pullout, which leads to
273
composite ductility.
AC C
EP
TE D
M AN U
SC
RI PT
244
274 275
4 CONCLUSIONS
276
High values of elastic modulus and hardness, 144 ± 48.9 GPa and 9 ± 2.5 GPa, respectively,
277
was found by means of nanoindentation imprints in the cellulose fiber, because of the
278
majority presence of C-S-H, portlandite and calcite phases. Then, cellulose fibers absorb
ACCEPTED MANUSCRIPT
water rich in cement hydration products, which induce the cellulose fiber stiffening by the
280
precipitation and re-precipitation of hydration products into the fiber, these phenomenon lead
281
to the so-called mineralization process. The increase of stiffness of the inner part of the
282
cellulose fibers is a drawback because can induce the loss of their reinforcement capacity. .
283
The present finding contributes to better understanding the interface properties between fiber
284
and matrix that affect cellulose fiber-cement materials performance, in terms of macroscopic
285
mechanical properties. It is difficult to obtain the effective elastic modulus these composites
286
because the low adhesion between the hydrophilic cellulose fibers and cement matrix,
287
mineralization of the cellulose fiber and consequently a significant difference of elastic
288
modulus between matrix, interface transition zone (ITZ), and cellulose fiber. These results
289
indicate a clear modification of the contribution of the cellulosic fiber in the stress distribution
290
and interfacial bond strength in the cement based composite. Therefore, it is still a challenge
291
to modify nano and micro structures of the ITZ in order to improve mechanical properties in
292
macro scale of the extruded fiber-cement composites.
293
M AN U
SC
RI PT
279
5 ACKNOWLEDGEMENTS
295
The authors acknowledge by financial support provided by Fundação de Amparo à Pesquisa
296
do Estado de São Paulo (FAPESP, process nº 2013/03823-8), Coordenação de
297
Aperfeiçoamento de Pessoal de Nível Superior (CAPES, process nº 3886/2014) and Conselho
298
Nacional de Desenvolvimento Científico e Tecnológico (CNPq), in Brazil. Special thanks for
299
Fibria and Infibra for providing raw materials the development of this work.
302 303 304 305 306 307 308 309 310 311 312 313
EP
301
AC C
300
TE D
294
ACCEPTED MANUSCRIPT
314
6 REFERENCES
315
[1]
316
in simulated cement microstructures, Cem. Concr. Compos. 80 (2017) 224–234.
317
doi:10.1016/j.cemconcomp.2017.03.008.
318
[2]
319
between cement paste and aggregate in concrete, Interface Sci. 12 (2004) 411–421.
320
doi:10.1023/B:INTS.0000042339.92990.4c.
321
[3]
322
lignocellulosic fiber and matrix in cement-based composites, Editors: Holmer Savastano
323
Junior, Juliano Fiorelli and Sergio Francisco dos Santos, Woodhead Publishing, 2017, pp. 27-
324
68. doi:10.1016/j.actamat.2015.02.029.
325
[4]
326
new high toughness polypropylene (PP) fibers in air-cured Hatschek process, Constr. Build.
327
Mater. 24 (2010) 171–180. doi:10.1016/j.conbuildmat.2009.06.019.
328
[5]
329
carbonation on cementitious roofing tiles reinforced with lignocellulosic fibre, Constr. Build.
330
Mater. 24 (2010) 193–201. doi:10.1016/j.conbuildmat.2007.11.018.
331
[6]
332
Eucalyptus pulp fibres as alternative reinforcement to engineered cement-based composites,
333
Ind. Crops Prod. 31 (2010) 225–232. doi:10.1016/j.indcrop.2009.10.009.
334
[7]
335
R.N. (Ed.), Nat. Fibre Reinf. Cem. Concr., Glasgow: Blackie,1988. pp.208-42. (Concrete
336
Technology and Design, 5), 1988.
337
[8]
338
Extruded Cement Based Composites Reinforced with Sugar Cane Bagasse Fibres, Key Eng.
339
Mater. 517 (2012) 450–457. doi:10.4028/www.scientific.net/KEM.517.450.
340
[9]
341
8 (1973) 352–362. doi:10.1007/BF00550155.
342
[10]
343
extrusion, Can. J. Civ. Eng. 27 (2000) 543–552. doi:10.1139/cjce-27-3-543.
344
[11]
345
composites, Cem. Concr. Compos. 21 (1999) 49–57. doi:10.1016/S0958-9465(98)00038-9.
346
[12]
347
cement components for roofing, Constr. Build. Mater. 13 (1999) 433–438.
348
doi:10.1016/S0950-0618(99)00046-X.
P. Carrara, L. De Lorenzis, Consistent identification of the interfacial transition zone
RI PT
K.L. Scrivener, A.K. Crumbie, P. Laugesen, The interfacial transition zone (ITZ)
SC
S.F. Santos, R.S. Teixeira, H. Savastano Jr., Interfacial transition zone between
M AN U
S. Ikai, J.R. Reichert, A. V. Rodrigues, V.A. Zampieri, Asbestos-free technology with
G.H.D. Tonoli, S.F. Santos, A.P. Joaquim, H. Savastano Jr, Effect of accelerated
TE D
G.H.D. Tonoli, H. Savastano, E. Fuente, C. Negro, A. Blanco, F.A. Rocco Lahr,
EP
R.S.P. COUTTS, Wood fibre reinforced cement composites., in: (Ed.) In: SWAMY,
AC C
R.S. Teixeira, G.H.D. Tonoli, S.F. Santos, J. Fiorelli, H. Savastano Jr., F.A.R. Lahr,
J. Aveston, A. Kelly, Theory of multiple fracture of fibrous composites, J. Mater. Sci.
Y. Shao, S. Moras, N. Ulkem, G. Kubes, Wood fibre - cement composites by
H. Savastano Jr, V. Agopyan, Transition zone studies of vegetable fibre-cement paste
H. Savastano Jr, V. Agopyan, A.M. Nolasco, L. Pimentel, Plant fibre reinforced
ACCEPTED MANUSCRIPT
349
[13]
W. Zhu, P.J.M. Bartos, Application of depth-sensing microindentation testing to study
350
of interfacial transition zone in reinforced concrete, Cem. Concr. Res. 30 (2000) 1299–1304.
351
doi:10.1016/S0008-8846(00)00322-7.
352
[14]
353
Cement-based Composites, J. Proc. 83 (1986) 597–605.
354
[15]
355
Fiber-Reinforced Cement Paste, Mater. J. 97 (2000) 162–167.
356
[16]
357
modulus using load and displacement sensing indentation experiments, Mater. Res. Soc. 7
358
(1992) 1564–1583.
359
[17]
360
prediction of macroscale elastic properties of high performance cementitious composites,
361
Cem. Concr. Compos. 45 (2014) 57–68. doi:10.1016/j.cemconcomp.2013.09.013.
362
[18]
363
Effect of fibre morphology on flocculation of fibre-cement suspensions, Cem. Concr. Res. 39
364
(2009) 1017–1022. doi:10.1016/j.cemconres.2009.07.010.
365
[19]
366
Standard Specification for Portland Cement.
367
[20]
368
study on vegetable wastes as reinforcement in extruded fibre-cement., in: Int. Conf. Non-
369
Conventional Mater. Technol. Ecol. Mater. Technol. Sustain. Constr., 2007, Maceió. Anais do
370
Brasil-NOCMAT. (Abmtenc) Maceió., 2007.
371
[21]
372
Supercritical carbonation treatment on extruded fibre–cement reinforced with vegetable
373
fibres, Cem. Concr. Compos. 56 (2015) 84–94. doi:10.1016/j.cemconcomp.2014.11.007.
374
[22]
375
Rocco Lahr, Cellulose modified fibres in cement based composites, Compos. Part A Appl.
376
Sci. Manuf. 40 (2009) 2046–2053. doi:10.1016/j.compositesa.2009.09.016.
377
[23]
378
Pescatori Silva, H. Savastano Jr., Effect of eucalyptus pulp refining on the performance and
379
durability of fibre-cement composites., J. Trop. For. Sci. 25 (2013) 400–409.
380
[24]
381
Dissolution and Carbonation of Portlandite [Ca(OH)2] Single Crystals, Environ. Sci. Technol.
382
47 (2013) 11342–11349. doi:10.1021/es402061c.
383
[25]
S. Wei, J.A. Mandel, S. Said, Study of the Interface Strength in Steel Fiber-Reinforced
RI PT
W.M. Cross, K.H. Sabnis, L. Kjerengtroen, J.J. Kellar, Microhardness Testing of
W. Oliver, G. Pharr, An improved technique for determining hardness and elastic
M AN U
SC
W.R.L. Da Silva, J. Němeček, P. Štemberk, Methodology for nanoindentation-assisted
G.H.D. Tonoli, E. Fuente, C. Monte, H. Savastano Jr, F.A. Rocco Lahr, A. Blanco,
ASTM C 150, American Society for Testing and Materials – ASTM C150/C150M-11.
TE D
Y.J.M. Soto, G.H.D. Tonoli, R.S. Teixeira, S.F. Santos, H. Savastano Jr, Prospective
AC C
EP
S.F. Santos, R. Schmidt, A.E.F.S. Almeida, G.H.D. Tonoli, H. Savastano Jr,
G.H.D. Tonoli, U.P. Rodrigues Filho, H. Savastano Jr, J. Bras, M.N. Belgacem, F.A.
G.H.D. Tonoli, S.F. Santos, R.S. Teixeira, M.A. Pereira-Da-Silva, F.A.R. Lahr, F.H.
E. Ruiz-Agudo, K. Kudłacz, C. V. Putnis, A. Putnis, C. Rodriguez-Navarro,
J.A. Larbi, J.M.J.M. Bijen, Interaction of polymers with portland cement during
ACCEPTED MANUSCRIPT
384
hydration. A study, Cem. Concr. Res. 20 (1990) 139–147. doi:10.1016/0008-8846(90)90124-
385
G.
386
[26]
387
polymer modification on the formation of high sulphoaluminate or ettringite - type (AFt)
388
crystals in polymer-modified mortars, Cem. Concr. Res. 24 (1994) 1492–1494.
389
doi:10.1016/0008-8846(94)90163-5.
390
[27]
H.F.W. Taylor, Cement chemistry, London: Academic press limited, 1990.
391
[28]
J.I. Escalante-Garcia, G. Mendoza, J.H. Sharp, Indirect determination of the Ca/Si
392
ratio of the C-S-H gel in Portland cements, Cem. Concr. Res. 29 (1999) 1999–2003.
393
doi:10.1016/S0008-8846(99)00191-X.
394
[29]
395
cycling on pulp fiber-cement composites, Cem. Concr. Res. 36 (2006) 1240–1251.
396
doi:10.1016/j.cemconres.2006.03.020.
397
[30]
398
nanoindentation testing to study of the interfacial transition zone in steel fiber reinforced
399
mortar, Cem. Concr. Res. 39 (2009) 701–715. doi:10.1016/j.cemconres.2009.05.002.
400
[31]
401
Adv. Cem. Based Mater. 4 (1996) 48–57. doi:10.1016/1065-7355(96)00035-1.
402
[32]
403
micromechanical properties of fibre reinforced cementitious composite., in: 6th RILEM
404
Symp. Fibre Reinf. Concr., BEFIB, 2004, Varenna, Italy, (2004) pp. 401–410., 2004: pp.
405
401–410.
406
[33]
407
In-built nanotechnology ,” Cerâmica. 55 (2009) 199–205. (in portuguese).
408
[34]
409
nanoindentation of elastic modulus and hardness of pure constituents of Portland cement
410
clinker, Cem. Concr. Res. 31 (2001) 555–561. doi:10.1016/S0008-8846(00)00505-6.
411
[35]
412
cement composites cured in a normal environment, Int. J. Cem. Compos. Light. Concr. 11
413
(1989) 99–109. doi:10.1016/0262-5075(89)90120-6.
414
[36]
415
sensitive sisal and coconut fibres in cement mortar composites, Cem. Concr. Compos. 22
416
(2000) 127–143. doi:10.1016/S0958-9465(99)00039-6.
417
[37]
418
composites to wet/dry cycling, Cem. Concr. Res. 35 (2005) 1646–1649.
RI PT
M.U.K. Afridi, Z.U. Chaudhary, Y. Ohama, K. Demura, M.Z. Iqbal, Effects of
M AN U
SC
B.J. Mohr, J.J. Biernacki, K.E. Kurtis, Microstructural and chemical effects of wet/dry
X.H. Wang, S. Jacobsen, J.Y. He, Z.L. Zhang, S.F. Lee, H.L. Lein, Application of
S. Igarashi, A. Bentur, S. Mindess, Microhardness testing of cementitious materials,
TE D
J. Meček, P. Kabele, Z. Bittnar, Nanoindentation based assessment of
EP
H.L. Rossetto, M.F. de Souza, V.C. Pandolfelli, Adesão em materiais cimentícios : “
AC C
K. Velez, S. Maximilien, D. Damidot, G. Fantozzi, F. Sorrentino, Determination by
A. Bentur, S.A.S. Akers, The microstructure and ageing of cellulose fibre reinforced
R.D. Tolêdo Filho, K. Scrivener, G.L. England, K. Ghavami, Durability of alkali-
B.J. Mohr, H. Nanko, K.E. Kurtis, Durability of thermomechanical pulp fiber-cement
ACCEPTED MANUSCRIPT
419
doi:10.1016/j.cemconres.2005.04.005.
420
[38]
421
to wet/dry cycling, Cem. Concr. Compos. 27 (2005) 435–448.
422
doi:10.1016/j.cemconcomp.2004.07.006.
423
[39]
424
and durability of cement based composites reinforced with refined sisal pulp, Mater. Manuf.
425
Process. 22 (2007) 149–156. doi:10.1080/10426910601062065.
B.J. Mohr, H. Nanko, K.E. Kurtis, Durability of kraft pulp fiber – cement composites
426 427 428
SC
429 430
M AN U
431 432 433 434 435 436
441 442 443 444 445 446 447 448 449 450 451 452 453
EP
440
AC C
439
TE D
437 438
RI PT
G.H.D. Tonoli, A.P. Joaquim, M.-A. Arsène, K. Bilba, H. Savastano Jr, Performance
454
ACCEPTED MANUSCRIPT
Table
455 456 457 458 459
Table 1. Physical characteristics of the unbleached eucalyptus pulp [18]. Table 2. Mix design used in the production of the composites. Table 3. Average and standard deviation values of elastic modulus and hardness obtained for
461
each structure of the composite by nanoindentation technique1
462 463 464
SC
465 466
M AN U
467 468 469 470 471 472
477 478 479 480 481 482 483 484 485 486 487 488 489
EP
476
AC C
475
TE D
473 474
RI PT
460
490
ACCEPTED MANUSCRIPT
Figure
491 492
Figure 1 – (a) Extruder machine used for molding; (b) side view showing the mass exit
493
through the die; and (c) details of the operating mechanism of the extrusion process (adapted
494
from [16]).
495 Figure 2 – Left: Scanning electron microscopy (SEM) micrographs in back scattering electron
497
(BSE) mode of the polished cross-section of an extruded composite. Right: Energy dispersive
498
spectra (EDS) referring to the fiber-matrix interface (upper) and to the bulk matrix (bottom).
RI PT
496
499
Figure 3 – AFM topography image of the interface fracture. Arrows show the possible
501
Ca(OH)2 with carbonated production in the interfacial zone between the cellulose fiber and
502
cement matrix.
M AN U
SC
500
503 504
Figure 4 – Scanning electron microscopy (SEM) image in backscattering electron (BSE)
505
mode of the polished cross-section in an extruded composite. Energy dispersive spectra
506
(EDS) refer to the composition of the anhydrous grain (spot A) and bulk matrix region (spot
507
B).
TE D
508 509
Figure 5 – (a) Light micrograph of the nanoindentation spots applied in line along the bulk
510
section of the composite; (b) individual values of elastic modulus for each spot of the
511
micrograph; and (c) individual values of hardness for each spot of the micrograph.
EP
512
Figure 6 – (a) Light micrograph of the nanoindentation spots applied in line along the
514
different structures (anhydrous grain, bulk density, cellulose fiber and interfacial/transition
515
zone between fiber and matrix) of the composite; (b) individual values of elastic modulus for
516
each spot of the micrograph; and (c) individual values of hardness for each spot of the
517
micrograph.
518
AC C
513
519
Figure 7 – (a) Scanning electron microscopy (SEM) micrographs in back scattering electron
520
(BSE) mode of the polished cross-section of an extruded composite; and (b) energy dispersive
521
spectra (EDS) of the spots 1 and 2 in the SEM micrographs.
522 523 524
525
ACCEPTED MANUSCRIPT
Table 1 Average length (mm)
Average width (µm)
Aspect Ratio
Fibrous material (106 fibers/g)
Content of fines by mass (%)
0.83 ± 0.05
16.4 ± 0.2
51
18.17 ± 1.11
25.7 ± 0.6
526 527
RI PT
528 529 530 531
SC
532 533 534
M AN U
535 536 537 538 539
543 544 545 546 547 548 549 550 551 552 553 554 555 556
EP
542
AC C
541
TE D
540
558 559
ACCEPTED MANUSCRIPT
Table 2 Raw material
Content (% by mass)
Portland cement (CPV-ARI) a
70
Limestone filler b
27
Unbleached eucalyptus pulp c
3
(a) ASTM C150 type I [19], provided by Caue, (b) Itaú, (c) Fibria S.A.
RI PT
557
560 561 562
SC
563 564 565
M AN U
566 567 568 569 570
574 575 576 577 578 579 580 581 582 583 584 585 586 587
EP
573
AC C
572
TE D
571
ACCEPTED MANUSCRIPT
588
590 591 592
Table 3 Structures
Elastic modulus (GPa)
Hardness (GPa)
Interfacial/transition zone
17.2 ± 5.6 c
0.4 ± 0.2 c
Bulk matrix
72.7 ± 14.3 b
5.3 ± 1.8 b
Anhydrous grain
89.9 ± 17.1 b
4.5 ± 1.9 b
Cellulosic fiber
144.3 ± 48.9 a
8.9 ± 2.5 a
1
Lowercase letters (a, b and c) in the same column represent comparisons between conditions of the composites. Different letters indicate significant differences amongst the properties of the composites.
593
SC
594 595
M AN U
596 597 598 599 600
605 606 607 608 609 610 611 612 613 614 615 616 617 618
EP
604
AC C
603
TE D
601 602
RI PT
589
ACCEPTED MANUSCRIPT
SC
RI PT
619
621
M AN U
620 Figure 1
622 623 624
628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
EP
627
AC C
626
TE D
625
ACCEPTED MANUSCRIPT
643
RI PT
Interface
SC
Bulk matrix
M AN U
644 645 646
Figure 2
647 648 649
653 654 655 656 657 658 659 660 661 662 663 664 665 666
EP
652
AC C
651
TE D
650
ACCEPTED MANUSCRIPT
RI PT
667
SC
5
µm 668 669 670
Figure 3
671 672
676 677 678 679 680 681 682 683 684 685 686 687 688 689
EP
675
AC C
674
TE D
673
M AN U
10
ACCEPTED MANUSCRIPT
690
Anhydrous grain
B
SC
Bulk matrix
RI PT
A
691 692 Figure 4
M AN U
693 694 695 696 697
701 702 703 704 705 706 707 708 709 710 711 712 713
EP
700
AC C
699
TE D
698
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
(a)
AC C
EP
TE D
(b)
714 715
(c) Figure 5
ACCEPTED MANUSCRIPT
Transition zone
81
Cellulosic fibre Anhydrous grain
SC
71
RI PT
101 91
Bulk matrix
(b)
AC C
EP
TE D
M AN U
(a)
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
(c) 716 717
Figure 6
718
722 723 724 725 726 727 728 729 730 731 732 733 734 735
EP
721
AC C
720
TE D
719
ACCEPTED MANUSCRIPT
736
1 Hydration products in the cellulosic fiber
1 Cellulosic fiber
SC
(a)
(b)
737 738
Anhydrous grain
M AN U
30 µm
Bulk matrix
Transition zone
RI PT
2
Figure 7
AC C
EP
TE D
739
2
S0958-9465(16)30464-4
DOI:
10.1016/j.cemconcomp.2017.09.018
Reference:
CECO 2916
To appear in:
Cement and Concrete Composites
Received Date: 19 August 2016 Revised Date:
14 July 2017
Accepted Date: 29 September 2017
Please cite this article as: R.S. Teixeira, G.H.D. Tonoli, S.F. Santos, E. Rayon, V. Amigó, H. Savastano Jr., F.A.R. Lahr, Nanoindentation study of the interfacial zone between cellulose fiber and cement matrix in extruded composites, Cement and Concrete Composites (2017), doi: 10.1016/ j.cemconcomp.2017.09.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Nanoindentation study of the interfacial zone between cellulose fiber and
2
cement matrix in extruded composites
3 4
R.S. Teixeira1; G.H.D. Tonoli2; S.F. Santos3, E. Rayon4, V. Amigó4; H. Savastano Jr5; F.A.
5
Rocco Lahr1. 1
University of São Paulo (USP), Department of Structural Engineering. School of Engineering of São Carlos, Brazil; 2 Federal University of Lavras (UFLA), Department of Forest Science, Brazil; 3 São Paulo State Universit (UNESP), School of Engineering, Department of Materials and Technology, Brazil; 4 Universitat Politècnica de València (UPV), Department of Mechanical and Materials Engineering, Spain. 5 University of São Paulo (USP), Department of Biosystems Engineering, Brazil.
15
ABSTRACT
16
The present study shows the application of the nanoindentation technique to evaluate the
17
properties of the cellulose fiber-cement matrix interfacial zone in composites prepared with an
18
auger extruder. The degree of strength of the bond between fiber and matrix is recognized as
19
important variable that influences macro-mechanical properties, such as modulus of rupture
20
and toughness of cement based composites. The nanoindentation measurements showed the
21
highest hardness and elastic modulus in the part inner of the cellulosic fiber after hydration
22
process due to precipitation and re-precipitation of cement hydration products. These results
23
indicate that mineralization of the cellulosic fibers can affect the stress distribution and
24
interfacial bond strength in the cement based composite.
TE D
M AN U
SC
RI PT
6 7 8 9 10 11 12 13 14
EP
25
KEYWORDS: vegetable fiber, microstructure, mechanical properties, ordinary Portland
27
cement, transition zone.
28
AC C
26
29
1 INTRODUCTION
30
It is widely accepted that concrete at the mesoscopic scale should be treated as a three-phase
31
composite material constituted by bulk hardened cement paste, aggregates and an interfacial
32
transition zone (ITZ) between these two. [1] Therefore, traditional ITZ is the region of the
33
cement matrix around the aggregate particles, which is perturbed by the presence of the
34
aggregate. Its origin lies in the packing of the cement grains against the much larger
35
aggregate, which leads to a local increase in porosity and predominance of smaller cement
36
particles in this region. The ITZ is region of gradual transition and is highly heterogeneous,
37
nevertheless the average microstructural features may be measured [2].
ACCEPTED MANUSCRIPT
The cementitious composites reinforced with cellulosic fibers are complex materials. Thus,
39
cement based composite design must be considered for the interrelation between interfacial
40
transition zone, fibers, matrix and processing method. Particle and fiber size distribution,
41
particles and fibers packing and rheological behavior of the cement paste with fibers in the
42
fresh state also interfere on the interfacial transition zone of the hardened cement based
43
composite. Furthermore, changes in the mechanical properties over time can occur due to
44
microstructural changes in the fiber-matrix interface and bulk, as a consequence of the
45
continued hydration process that occurs in the fiber surroundings, i.e. if the interfacial
46
transition zone becomes denser due to re-precipitation of cement hydration products in the
47
porous layers. [3]. Therefore, in this manuscript an interfacial transition zone (ITZ)
48
terminology is related to a region between bulk hardened cement paste and cellulosic fiber.
49
Kraft pulp fiber-reinforced cement-based materials have being increasingly used in most
50
developing countries as construction materials [4–6]. Fibers usually increase mechanical
51
strength and the capacity of cement based composite to absorb energy through the distribution
52
of microcracks into the material. The bonding between cellulose fiber and Portland cement
53
plays a major role in the transference of the load from the matrix to the reinforcement,
54
impacting the mechanical behavior of the resulting fiber-cement material. This bonding can
55
be of physical (mechanical interlocking) or chemical (mainly hydrogen bonds) nature, or a
56
combination of both [7].
57
Extrusion process has a potential of producing fiber reinforced cementitious composites with
58
higher performance compared to the traditional slurry dewatering processes such as Hatschek
59
and Flow-on methods. Several advantages of extrusion moulding can be pointed for
60
manufacturing of fibrous composites: (i) low implementation cost; (ii) low capillary porosity
61
of extruded composites; and (iii) low water/cement ratio, which plays a significant role for
62
improving of the fiber-matrix interface zone [8–10]. As mentioned above, the control of fiber-
63
matrix interface zone in extruded composites depends on cellulosic fibers and matrix
64
constituent’s properties, cure procedure as well as weathering conditions that the composite
65
are exposed. Previous studies highlighted that water/cement ratio influences the interface zone
66
thickness and that lower porosity may increase composite strength and matrix toughness [9–
67
12]. Therefore, the optimization of the interfacial zone is not an easy task since it is necessary
68
to satisfy the equilibrium between fiber-matrix adhesion and fiber pull-out. In addition, there
69
is still a lack of information about the consequences of the different types of interface zone
70
between cellulose fibers and matrix, on composite mechanical properties. Thus, an effort is
71
required in order to understand the mechanical properties at fiber-matrix transition zone.
AC C
EP
TE D
M AN U
SC
RI PT
38
ACCEPTED MANUSCRIPT
The macroscale mechanical properties of bulk materials are dependent on the properties of
73
their microscale/nanoscale constituents and the interface between these constituents.
74
Microindentation technique is one of the developing tests that have been used for evaluating
75
mechanical characteristics of microsized zones in diverse cement based composites [13–15].
76
Nevertheless, microindentation is not capable to detect mechanical differences at the
77
interfaces between cement matrix and reinforced fibers, due to some limitations, such as a
78
resolution of microindentation system in terms of load is about 5.0 mN, to analyze some
79
structural features because of the heterogenity of the specimen, inadequate spacing between
80
indents or form edges, operator’s visual acuity. Quasistatic nanoindentation has become the
81
standard technique used for nanomechanical characterization of small volumes of material,
82
such as elastic modulus, hardness at the nanoscale. It is used machines that can record small
83
load and displacement with high accuracy and precision, during loading and unloading can be
84
continuously monitored with high resolution in the nanonewton and nanometer ranges. In this
85
technique is used usually the Berkovich tip is a 3-sided pyramid that have an apex appropriate
86
for investigating the interface transition zone between fiber and matrix. Therefore,
87
nanoindentation technique can contribute to quantitatively measure the underlying source of
88
the bulk response at the nanoscale in cement based composites that are heterogeneous
89
material from both a chemical composition and nanoscale structure standpoint. Being able to
90
evaluate at this scale allows micro and nano structural "tuning" to be performed to optimize
91
bulk properties [16,17]. The aim of this study was to evaluate the interfacial zone between
92
fibers to cement matrix in extruded composites reinforced with eucalyptus pulp, using a
93
nanoindentation technique allied with microscopic examinations.
94
EP
TE D
M AN U
SC
RI PT
72
2 EXPERIMENTAL
96
2.1 Materials
97
Unbleached unrefined eucalyptus (Eucalyptus grandis) Kraft pulp was provided by Fibria
98
S/A, Brazil. The cellulose pulp was collected directly from the mill, prior to drying and
99
pressing. It was extensively washed with water and centrifuged to remove any residual
AC C
95
100
chemicals from the pulping processes.
101
Eucalyptus pulps were analyzed about morphology (length and width) and concentration of
102
particles (fibers per gram and content of fines). A fine element (small particles with length
103
lower than 200 µm that are commonly called fines, which also contain band-like materials
104
from both primary wall and secondary wall layers of the fibers) was detected in the pulp with
105
dimensions lower than those of fibers. Fines do not contribute significantly to cement based
ACCEPTED MANUSCRIPT
106
composite strength but act more as filler [18]. Physical characteristics of the unbleached
107
eucalyptus pulp are showed in Table 1.
108 2.2 Composite mix design
110
Formulation used in the composite production was based on previous study [8], and is
111
described in Table 2.
112
Portland cement and limestone particles distribution was performed by laser particle size
113
analyzer (Malvern Mastersizer S long bed, version 2.19). Cement and limestone particles
114
showed 50% of their size below 11 µm and 16 µm respectively.
115
Hydroxypropylmethylcellulose (HPMC) from Aditex, is a superplasticizer which disperses
116
cement particles, reducing the flow stress and the viscosity of the cement paste. Polyether
117
carboxylic provided by Grace, commercially called ADVA 170, is an additive which reduces
118
the surface tension of the water and thus causes an increase in the surface area of the liquid
119
which improves the water absorption, forming a lubrication layer between the wall of the
120
extruder and the surface of the fiber cement board. HPMC with 86,000 average molecular
121
weight and 5.39 cPs viscosity (at 2% concentration in water at 20°C) and polyether carboxylic
122
commercially named ADVA 170, were used as rheological modifiers in fiber-cement, with
123
dosage of 1 wt% of the total mass of the particulate inorganic materials (Portland cement and
124
limestone) to promote the pseudo-plastic behavior of the composite, which, in turn, enabled
125
the extrusion process.
126
TE D
M AN U
SC
RI PT
109
2.3 Composite extrusion
128
The mass was mixed in a mechanical Eirich intensive mixer (capacity of 10 L) for 5 min at
129
125 rpm, 5 min at medium velocity (220 rpm) and finally 5 min at high velocity (450 rpm).
130
The water-cement (w/c) ratio was 0.33. The composite mixture was transferred to a Verdes
131
051 extruder (Figure 1a-b). The speed of the extruder was 4 mm/s; die width/diameter ratio
132
was 3.3. Composites were molded in 15 mm x 50 mm x 200 mm plates. The illustration of the
133
chamber parts of the extruder is presented in Figure 1c.
134
The composite mixture was re-circulated into the extruder for 5 min before tailoring the
135
samples. Pads with 15 mm x 50 mm x 200 mm were cured at water vapor saturated
136
environment (in sealed plastic bags) at 25 ± 2ºC for two days. Subsequently, the specimens
137
were maintained in water vapor saturated environment (in sealed plastic bags) and placed in a
138
chamber at 45ºC for five days (thermal curing) totalizing 7 days of cure [21].
139 140
AC C
EP
127
ACCEPTED MANUSCRIPT
2.4 Microstructure characterization
142
The specimens were cut out from the middle section of a larger volume of fiber cement, with
143
minimal deformation, by using a precision cutter, Buehler, model Isomet, with precision
144
diamond saw, 15LC, and using an isopropyl alcohol based cutting fluid to cut. This method
145
provides roughness smaller. The specimens were embedded in epoxy resin (MC-
146
DUR1264FF) previously to be visualized in a scanning electron microscope (SEM). A final
147
polishing was carried out using sequentially 8-4 µm, 4-2 µm and 1-0 µm diamond abrasives,
148
as reported elsewhere [22,23]. The polished samples were carbon coated before being
149
analyzed in a JEOL JSM6300 electron microscope with acceleration voltage of 20 kV. A
150
back-scattered electron (BSE) detector was used to observe the morphological features on the
151
cross-section and polished surfaces. BSE imaging permits the identification of cementitious
152
phases since the electron scattering change. Dark and light areas are related to lighter and
153
heavier elements respectively. Energy dispersive spectrometry (EDS) analysis was performed
154
in order to identify the chemical composition in different spots on the same polished surface
155
specimens.
M AN U
SC
RI PT
141
156
2.4.1 Atomic force microscopy characterization (AFM)
158
In contrast to ductile material, cement based composite have regions with higher hardness,
159
lower ductility, and a basically brittle nature. Due to the brittleness, cellulosic fibers and
160
porosity of the composite it is quite often difficult to cut and polish. All areas of the
161
specimens observed by nanoindentation were cut and polished with uniformity and carefully
162
to avoid any pullout of fibers, but as the composite is heterogeneous the response of each
163
region was different according to its resistance. Specimens for AFM characterization were cut
164
with same system mentioned above and embedded in polyacrylamide. The semi-automatic
165
polisher was used to adjust polishing velocity and pressure, i.e. all regions were submitted to
166
same conditions of preparation. Then, the dried surfaces were grinded in a sequence of SiC
167
papers (grit #500, #1000, #2000 and #4000) and polished in a cloth sequentially with 9 µm,
168
6 µm and 3 µm diamond sprays. Finally, the specimen surfaces were cleaned by washing in
169
alcohol and ultrasonic bathing.
170
The characterization of the fiber-matrix interface zone was carried out using the tapping-mode
171
atomic force microscopy (TM-AFM) to obtain height and phase imaging data simultaneously,
172
with Nanoscope III-Digital Instruments. Silicon cantilever with spring constant of around
173
70 N/m and a scan area of 15 µm x 15 µm was used. Samples were prepared in the laboratory
174
environment with temperature around 25°C and relative humidity between 50% and 65%.
AC C
EP
TE D
157
ACCEPTED MANUSCRIPT
175
AFM images were obtained in the same laboratory environment conditions. The test locations
176
were guided by means of an incorporated optical light microscope.
177 2.5 Nanoindentation tests
179
Specimens for nanoindentation tests were prepared the same methodology that specimens to
180
AFM characterizations. Nanoindentation was performed in an Agilent G-200 instrument
181
working in continuous stiffness measurement (CSM), configured at 2 nm of harmonic
182
amplitude, 45 Hz frequency, and using a Berkovich indenter. The test locations were guided
183
by means of an incorporated optical light microscope. To rule out the influence of surface
184
morphology, the indentation depth should be much greater than the characteristic size of
185
surface roughness to avoid errors and size effect in the calculations [17]. Tests were
186
programmed to reach from 100 to 300 nm-indented depths until the maximum depth of 1000
187
nm. Therefore, the depth of the imprint spots used in the nanoindentation was to minimize
188
surface roughness of the samples. Analyses were performed with 8 nm distance between the
189
measurement points [17]. It was performed more than 20 analyses for each region of the
190
composite (bulk matrix, interface or transition zone, anhydrous grain and cellulose fibers).
191
To guarantee a significance results for nanoindentation tests were choose under a completely
192
randomized design the ITZ areas to analyze a linear sequence with four regions (transition
193
zone, cellulose fiber, transition zone, bulk matrix and anhydrous grain or follow sequence:
194
anhydrous grain, bulk matrix, cellulose fiber, transition zone) and twenty replications. The
195
effects of hardness and elastic modulus were compared via statistical contrast by student test
196
at 5% of significance [17]. The ACTION statistical program was used. Therefore, the
197
interfaces towards fibers and matrix were studied in greater detail.
SC
M AN U
TE D
EP
198
RI PT
178
3 RESULTS AND DISCUSSION
200
3.1 Chemical and microstructure characterization
201
Figure 2 shows SEM-BSE micrographs of the extruded composites. Energy dispersive X-ray
202
spectroscopy (EDS) analysis (Figure 2a) indicate that fiber-matrix interfacial zone presents
203
high Ca/Si ratio (between 4 to 5), indicating the possible correspondence with Ca(OH)2
204
crystals with carbonated products, such as calcite, as corroborated by AFM analysis (Figure
205
3) and according to Ruiz-Agudo et al. [24]. This interfacial zone rich in Ca(OH)2 crystals has
206
been widely reported in literature for cementitious composites [25–29]. Bulk matrix revealed
207
an expected lower Ca/Si ratio (approximately 2.5) as observed in Figure 2b. The phase
208
composition of the anhydrous grains and matrix are shown in Figures 4a and 4b, respectively.
209
Spot B presents higher concentration of SiO2 and CaO due to the hydration products of
AC C
199
210
ACCEPTED MANUSCRIPT
cement in relation to the spot A (region of the anhydrous grain).
211 3.2 Mechanical characterization by nanoindentation
213
Figures 5a and 6a present the micrographs of the nanoindentation imprints done along the
214
bulk section of the composites. The spots numbered between 1-10 (Figure 5a), 25-61 (Figure
215
5a) and 81-90 (Figure 6a) represent the bulk matrix of the composite. The spots 15-16 (Figure
216
5a), 20-22 (Figure 5a) and 101-110 (Figure 6a) represents the interfacial zone. The spots 16-
217
19 (Figure 5a) and 91-100 (Figure 6a) represents the cellulosic fiber, while 62-70 (Figure 5a)
218
and 81-90 (Figure 6a) are spots related to anhydrous grains. The individual values of elastic
219
modulus (E) acquired for each spot are summarized in Figures 5b and 6b; while the individual
220
values of hardness (H) is presented in Figures 5c and 6c.
221
Table 3 shows the average and standard deviation values of elastic modulus and hardness
222
obtained from nanoindentation in the different regions studied. The indent spots in the bulk
223
matrix led to elastic modulus of around 70 GPa and hardness of around 5 GPa. These results
224
are similar to those found by Silva et al. [17], who studied nanoindentation in concrete
225
specimens. The bulk matrix of composites based on Portland cement normally presents
226
calcium silicate hydrates (C-S-H) and calcite phases as main components, which have higher
227
rigidity and relative lower porosity than other phases such as Ca(OH)2 and ettringite that are
228
formed by poorly packed structures. This is why the spots correspondent to the
229
interfacial/transition zone presents lower elastic modulus (~17 GPa) and hardness (~0.4 GPa).
230
These results of the elastic modulus and hardness vary in relation to the distance of the
231
transition zone, as showed in Figure 5. Similar result was found by Wang et al. [30] that
232
studied nanoindentation technique in concrete reinforced with steel fibers. The values of
233
hardness in the 11-12, 23-26 indent spots near to the interfacial zone are smaller than those in
234
the matrix region, which are also related to the higher porosity in the interfacial zone,
235
resultant of the low adhesion between fiber and matrix [31,32].
236
Other important observation from this study in nanoscale level is related to the hydration of
237
the cement grains in the composite. The incorporation of water during the mixing of the
238
composite constituents leads to cement hydrate grains that have much lower average atomic
239
numbers than cement grains that were not completely hydrated (anhydrous grains). Then,
240
strong contrast is obtained between unreacted (anhydrous) and reacted material (hydrates)
241
[33]. Anhydrous grain shown in Table 3 presents elastic modulus of around 90.0 GPa and
242
hardness of around 4.5 GPa, which was not significantly different from average values of
243
hydrated bulk matrix. Velez et al. [34] also studied nanoindentation in cement and did not
AC C
EP
TE D
M AN U
SC
RI PT
212
ACCEPTED MANUSCRIPT
find differences between hardness and elastic modulus in anhydrous phases such as alite
245
(C3S), calcium aluminate (C3A) and belite (C2S).
246
The highly alkaline pore water within the fiber-cement interface can induce the cellulose
247
fibers stiffening. It occurs because cellulose fibers are hydrophilic and absorb water rich in
248
cement hydration products, leading to the mineralization phenomenon proposed elsewhere
249
[35–37]. This phenomenon is related to hydrophilic nature of the cellulose fibers. Therefore,
250
this phenomenon is related to the hydrophilic nature of the cellulose fibers. This nature
251
depends on surface nature, surface porosity and lumens the inner of the cellulose fiber. The
252
internal humidity in the fiber can promote a continued hydration process, with precipitation
253
and re-precipitation internally, such as C-S-H, portlandite and calcites, as suggest by Figures
254
7a and 7b. Figure 7a shows hydration products within the cell wall of cellulosic fibers, while
255
the EDS spectra (Figure 7b) in the spots 1 (upper) and 2 (bottom) proves the large amounts of
256
Si and Ca into the cellulose fibers. It explains the higher values of elastic modulus (~144 GPa)
257
and hardness (~9 GPa) observed for the nanoindentation spots related to the cellulosic fiber
258
regions. Therefore, the average elastic modulus of mineralized cellulose fibers can reach
259
higher values than cement bulk matrix. The increase of stiffness of the cellulose fibers is a
260
drawback because can induce the loss of their reinforcement capacity. The efficiency of stress
261
transfer from matrix to fiber that is of fundamental importance in determining the mechanical
262
properties of short fibre reinforced composite is prejudiced. This result explain the premature
263
loss of ductility/toughness and increase of stiffness of composites reinforced with cellulose
264
fibers widely reported in literature [22,23,38,39].
265
Therefore, the fiber to matrix bonding is impaired because cellulose fiber tends to exchange
266
water with the matrix system by capillary action, leading to high dimensional variation and
267
detachment of the fiber surface from the matrix, as showed in Figure 7a, creating voids that
268
explain the low results of elastic modulus and hardness in the transition zone. Finally, the
269
present observations may state that an optimal situation would be the cellulose fibers
270
protection from excessive water absorption, what can be achieved by less hydrophilic surfaces
271
(either natural or after an adequate treatment) and without losing the fiber bridging. Fiber
272
bridging is important to guarantee the expend of energy during fiber pullout, which leads to
273
composite ductility.
AC C
EP
TE D
M AN U
SC
RI PT
244
274 275
4 CONCLUSIONS
276
High values of elastic modulus and hardness, 144 ± 48.9 GPa and 9 ± 2.5 GPa, respectively,
277
was found by means of nanoindentation imprints in the cellulose fiber, because of the
278
majority presence of C-S-H, portlandite and calcite phases. Then, cellulose fibers absorb
ACCEPTED MANUSCRIPT
water rich in cement hydration products, which induce the cellulose fiber stiffening by the
280
precipitation and re-precipitation of hydration products into the fiber, these phenomenon lead
281
to the so-called mineralization process. The increase of stiffness of the inner part of the
282
cellulose fibers is a drawback because can induce the loss of their reinforcement capacity. .
283
The present finding contributes to better understanding the interface properties between fiber
284
and matrix that affect cellulose fiber-cement materials performance, in terms of macroscopic
285
mechanical properties. It is difficult to obtain the effective elastic modulus these composites
286
because the low adhesion between the hydrophilic cellulose fibers and cement matrix,
287
mineralization of the cellulose fiber and consequently a significant difference of elastic
288
modulus between matrix, interface transition zone (ITZ), and cellulose fiber. These results
289
indicate a clear modification of the contribution of the cellulosic fiber in the stress distribution
290
and interfacial bond strength in the cement based composite. Therefore, it is still a challenge
291
to modify nano and micro structures of the ITZ in order to improve mechanical properties in
292
macro scale of the extruded fiber-cement composites.
293
M AN U
SC
RI PT
279
5 ACKNOWLEDGEMENTS
295
The authors acknowledge by financial support provided by Fundação de Amparo à Pesquisa
296
do Estado de São Paulo (FAPESP, process nº 2013/03823-8), Coordenação de
297
Aperfeiçoamento de Pessoal de Nível Superior (CAPES, process nº 3886/2014) and Conselho
298
Nacional de Desenvolvimento Científico e Tecnológico (CNPq), in Brazil. Special thanks for
299
Fibria and Infibra for providing raw materials the development of this work.
302 303 304 305 306 307 308 309 310 311 312 313
EP
301
AC C
300
TE D
294
ACCEPTED MANUSCRIPT
314
6 REFERENCES
315
[1]
316
in simulated cement microstructures, Cem. Concr. Compos. 80 (2017) 224–234.
317
doi:10.1016/j.cemconcomp.2017.03.008.
318
[2]
319
between cement paste and aggregate in concrete, Interface Sci. 12 (2004) 411–421.
320
doi:10.1023/B:INTS.0000042339.92990.4c.
321
[3]
322
lignocellulosic fiber and matrix in cement-based composites, Editors: Holmer Savastano
323
Junior, Juliano Fiorelli and Sergio Francisco dos Santos, Woodhead Publishing, 2017, pp. 27-
324
68. doi:10.1016/j.actamat.2015.02.029.
325
[4]
326
new high toughness polypropylene (PP) fibers in air-cured Hatschek process, Constr. Build.
327
Mater. 24 (2010) 171–180. doi:10.1016/j.conbuildmat.2009.06.019.
328
[5]
329
carbonation on cementitious roofing tiles reinforced with lignocellulosic fibre, Constr. Build.
330
Mater. 24 (2010) 193–201. doi:10.1016/j.conbuildmat.2007.11.018.
331
[6]
332
Eucalyptus pulp fibres as alternative reinforcement to engineered cement-based composites,
333
Ind. Crops Prod. 31 (2010) 225–232. doi:10.1016/j.indcrop.2009.10.009.
334
[7]
335
R.N. (Ed.), Nat. Fibre Reinf. Cem. Concr., Glasgow: Blackie,1988. pp.208-42. (Concrete
336
Technology and Design, 5), 1988.
337
[8]
338
Extruded Cement Based Composites Reinforced with Sugar Cane Bagasse Fibres, Key Eng.
339
Mater. 517 (2012) 450–457. doi:10.4028/www.scientific.net/KEM.517.450.
340
[9]
341
8 (1973) 352–362. doi:10.1007/BF00550155.
342
[10]
343
extrusion, Can. J. Civ. Eng. 27 (2000) 543–552. doi:10.1139/cjce-27-3-543.
344
[11]
345
composites, Cem. Concr. Compos. 21 (1999) 49–57. doi:10.1016/S0958-9465(98)00038-9.
346
[12]
347
cement components for roofing, Constr. Build. Mater. 13 (1999) 433–438.
348
doi:10.1016/S0950-0618(99)00046-X.
P. Carrara, L. De Lorenzis, Consistent identification of the interfacial transition zone
RI PT
K.L. Scrivener, A.K. Crumbie, P. Laugesen, The interfacial transition zone (ITZ)
SC
S.F. Santos, R.S. Teixeira, H. Savastano Jr., Interfacial transition zone between
M AN U
S. Ikai, J.R. Reichert, A. V. Rodrigues, V.A. Zampieri, Asbestos-free technology with
G.H.D. Tonoli, S.F. Santos, A.P. Joaquim, H. Savastano Jr, Effect of accelerated
TE D
G.H.D. Tonoli, H. Savastano, E. Fuente, C. Negro, A. Blanco, F.A. Rocco Lahr,
EP
R.S.P. COUTTS, Wood fibre reinforced cement composites., in: (Ed.) In: SWAMY,
AC C
R.S. Teixeira, G.H.D. Tonoli, S.F. Santos, J. Fiorelli, H. Savastano Jr., F.A.R. Lahr,
J. Aveston, A. Kelly, Theory of multiple fracture of fibrous composites, J. Mater. Sci.
Y. Shao, S. Moras, N. Ulkem, G. Kubes, Wood fibre - cement composites by
H. Savastano Jr, V. Agopyan, Transition zone studies of vegetable fibre-cement paste
H. Savastano Jr, V. Agopyan, A.M. Nolasco, L. Pimentel, Plant fibre reinforced
ACCEPTED MANUSCRIPT
349
[13]
W. Zhu, P.J.M. Bartos, Application of depth-sensing microindentation testing to study
350
of interfacial transition zone in reinforced concrete, Cem. Concr. Res. 30 (2000) 1299–1304.
351
doi:10.1016/S0008-8846(00)00322-7.
352
[14]
353
Cement-based Composites, J. Proc. 83 (1986) 597–605.
354
[15]
355
Fiber-Reinforced Cement Paste, Mater. J. 97 (2000) 162–167.
356
[16]
357
modulus using load and displacement sensing indentation experiments, Mater. Res. Soc. 7
358
(1992) 1564–1583.
359
[17]
360
prediction of macroscale elastic properties of high performance cementitious composites,
361
Cem. Concr. Compos. 45 (2014) 57–68. doi:10.1016/j.cemconcomp.2013.09.013.
362
[18]
363
Effect of fibre morphology on flocculation of fibre-cement suspensions, Cem. Concr. Res. 39
364
(2009) 1017–1022. doi:10.1016/j.cemconres.2009.07.010.
365
[19]
366
Standard Specification for Portland Cement.
367
[20]
368
study on vegetable wastes as reinforcement in extruded fibre-cement., in: Int. Conf. Non-
369
Conventional Mater. Technol. Ecol. Mater. Technol. Sustain. Constr., 2007, Maceió. Anais do
370
Brasil-NOCMAT. (Abmtenc) Maceió., 2007.
371
[21]
372
Supercritical carbonation treatment on extruded fibre–cement reinforced with vegetable
373
fibres, Cem. Concr. Compos. 56 (2015) 84–94. doi:10.1016/j.cemconcomp.2014.11.007.
374
[22]
375
Rocco Lahr, Cellulose modified fibres in cement based composites, Compos. Part A Appl.
376
Sci. Manuf. 40 (2009) 2046–2053. doi:10.1016/j.compositesa.2009.09.016.
377
[23]
378
Pescatori Silva, H. Savastano Jr., Effect of eucalyptus pulp refining on the performance and
379
durability of fibre-cement composites., J. Trop. For. Sci. 25 (2013) 400–409.
380
[24]
381
Dissolution and Carbonation of Portlandite [Ca(OH)2] Single Crystals, Environ. Sci. Technol.
382
47 (2013) 11342–11349. doi:10.1021/es402061c.
383
[25]
S. Wei, J.A. Mandel, S. Said, Study of the Interface Strength in Steel Fiber-Reinforced
RI PT
W.M. Cross, K.H. Sabnis, L. Kjerengtroen, J.J. Kellar, Microhardness Testing of
W. Oliver, G. Pharr, An improved technique for determining hardness and elastic
M AN U
SC
W.R.L. Da Silva, J. Němeček, P. Štemberk, Methodology for nanoindentation-assisted
G.H.D. Tonoli, E. Fuente, C. Monte, H. Savastano Jr, F.A. Rocco Lahr, A. Blanco,
ASTM C 150, American Society for Testing and Materials – ASTM C150/C150M-11.
TE D
Y.J.M. Soto, G.H.D. Tonoli, R.S. Teixeira, S.F. Santos, H. Savastano Jr, Prospective
AC C
EP
S.F. Santos, R. Schmidt, A.E.F.S. Almeida, G.H.D. Tonoli, H. Savastano Jr,
G.H.D. Tonoli, U.P. Rodrigues Filho, H. Savastano Jr, J. Bras, M.N. Belgacem, F.A.
G.H.D. Tonoli, S.F. Santos, R.S. Teixeira, M.A. Pereira-Da-Silva, F.A.R. Lahr, F.H.
E. Ruiz-Agudo, K. Kudłacz, C. V. Putnis, A. Putnis, C. Rodriguez-Navarro,
J.A. Larbi, J.M.J.M. Bijen, Interaction of polymers with portland cement during
ACCEPTED MANUSCRIPT
384
hydration. A study, Cem. Concr. Res. 20 (1990) 139–147. doi:10.1016/0008-8846(90)90124-
385
G.
386
[26]
387
polymer modification on the formation of high sulphoaluminate or ettringite - type (AFt)
388
crystals in polymer-modified mortars, Cem. Concr. Res. 24 (1994) 1492–1494.
389
doi:10.1016/0008-8846(94)90163-5.
390
[27]
H.F.W. Taylor, Cement chemistry, London: Academic press limited, 1990.
391
[28]
J.I. Escalante-Garcia, G. Mendoza, J.H. Sharp, Indirect determination of the Ca/Si
392
ratio of the C-S-H gel in Portland cements, Cem. Concr. Res. 29 (1999) 1999–2003.
393
doi:10.1016/S0008-8846(99)00191-X.
394
[29]
395
cycling on pulp fiber-cement composites, Cem. Concr. Res. 36 (2006) 1240–1251.
396
doi:10.1016/j.cemconres.2006.03.020.
397
[30]
398
nanoindentation testing to study of the interfacial transition zone in steel fiber reinforced
399
mortar, Cem. Concr. Res. 39 (2009) 701–715. doi:10.1016/j.cemconres.2009.05.002.
400
[31]
401
Adv. Cem. Based Mater. 4 (1996) 48–57. doi:10.1016/1065-7355(96)00035-1.
402
[32]
403
micromechanical properties of fibre reinforced cementitious composite., in: 6th RILEM
404
Symp. Fibre Reinf. Concr., BEFIB, 2004, Varenna, Italy, (2004) pp. 401–410., 2004: pp.
405
401–410.
406
[33]
407
In-built nanotechnology ,” Cerâmica. 55 (2009) 199–205. (in portuguese).
408
[34]
409
nanoindentation of elastic modulus and hardness of pure constituents of Portland cement
410
clinker, Cem. Concr. Res. 31 (2001) 555–561. doi:10.1016/S0008-8846(00)00505-6.
411
[35]
412
cement composites cured in a normal environment, Int. J. Cem. Compos. Light. Concr. 11
413
(1989) 99–109. doi:10.1016/0262-5075(89)90120-6.
414
[36]
415
sensitive sisal and coconut fibres in cement mortar composites, Cem. Concr. Compos. 22
416
(2000) 127–143. doi:10.1016/S0958-9465(99)00039-6.
417
[37]
418
composites to wet/dry cycling, Cem. Concr. Res. 35 (2005) 1646–1649.
RI PT
M.U.K. Afridi, Z.U. Chaudhary, Y. Ohama, K. Demura, M.Z. Iqbal, Effects of
M AN U
SC
B.J. Mohr, J.J. Biernacki, K.E. Kurtis, Microstructural and chemical effects of wet/dry
X.H. Wang, S. Jacobsen, J.Y. He, Z.L. Zhang, S.F. Lee, H.L. Lein, Application of
S. Igarashi, A. Bentur, S. Mindess, Microhardness testing of cementitious materials,
TE D
J. Meček, P. Kabele, Z. Bittnar, Nanoindentation based assessment of
EP
H.L. Rossetto, M.F. de Souza, V.C. Pandolfelli, Adesão em materiais cimentícios : “
AC C
K. Velez, S. Maximilien, D. Damidot, G. Fantozzi, F. Sorrentino, Determination by
A. Bentur, S.A.S. Akers, The microstructure and ageing of cellulose fibre reinforced
R.D. Tolêdo Filho, K. Scrivener, G.L. England, K. Ghavami, Durability of alkali-
B.J. Mohr, H. Nanko, K.E. Kurtis, Durability of thermomechanical pulp fiber-cement
ACCEPTED MANUSCRIPT
419
doi:10.1016/j.cemconres.2005.04.005.
420
[38]
421
to wet/dry cycling, Cem. Concr. Compos. 27 (2005) 435–448.
422
doi:10.1016/j.cemconcomp.2004.07.006.
423
[39]
424
and durability of cement based composites reinforced with refined sisal pulp, Mater. Manuf.
425
Process. 22 (2007) 149–156. doi:10.1080/10426910601062065.
B.J. Mohr, H. Nanko, K.E. Kurtis, Durability of kraft pulp fiber – cement composites
426 427 428
SC
429 430
M AN U
431 432 433 434 435 436
441 442 443 444 445 446 447 448 449 450 451 452 453
EP
440
AC C
439
TE D
437 438
RI PT
G.H.D. Tonoli, A.P. Joaquim, M.-A. Arsène, K. Bilba, H. Savastano Jr, Performance
454
ACCEPTED MANUSCRIPT
Table
455 456 457 458 459
Table 1. Physical characteristics of the unbleached eucalyptus pulp [18]. Table 2. Mix design used in the production of the composites. Table 3. Average and standard deviation values of elastic modulus and hardness obtained for
461
each structure of the composite by nanoindentation technique1
462 463 464
SC
465 466
M AN U
467 468 469 470 471 472
477 478 479 480 481 482 483 484 485 486 487 488 489
EP
476
AC C
475
TE D
473 474
RI PT
460
490
ACCEPTED MANUSCRIPT
Figure
491 492
Figure 1 – (a) Extruder machine used for molding; (b) side view showing the mass exit
493
through the die; and (c) details of the operating mechanism of the extrusion process (adapted
494
from [16]).
495 Figure 2 – Left: Scanning electron microscopy (SEM) micrographs in back scattering electron
497
(BSE) mode of the polished cross-section of an extruded composite. Right: Energy dispersive
498
spectra (EDS) referring to the fiber-matrix interface (upper) and to the bulk matrix (bottom).
RI PT
496
499
Figure 3 – AFM topography image of the interface fracture. Arrows show the possible
501
Ca(OH)2 with carbonated production in the interfacial zone between the cellulose fiber and
502
cement matrix.
M AN U
SC
500
503 504
Figure 4 – Scanning electron microscopy (SEM) image in backscattering electron (BSE)
505
mode of the polished cross-section in an extruded composite. Energy dispersive spectra
506
(EDS) refer to the composition of the anhydrous grain (spot A) and bulk matrix region (spot
507
B).
TE D
508 509
Figure 5 – (a) Light micrograph of the nanoindentation spots applied in line along the bulk
510
section of the composite; (b) individual values of elastic modulus for each spot of the
511
micrograph; and (c) individual values of hardness for each spot of the micrograph.
EP
512
Figure 6 – (a) Light micrograph of the nanoindentation spots applied in line along the
514
different structures (anhydrous grain, bulk density, cellulose fiber and interfacial/transition
515
zone between fiber and matrix) of the composite; (b) individual values of elastic modulus for
516
each spot of the micrograph; and (c) individual values of hardness for each spot of the
517
micrograph.
518
AC C
513
519
Figure 7 – (a) Scanning electron microscopy (SEM) micrographs in back scattering electron
520
(BSE) mode of the polished cross-section of an extruded composite; and (b) energy dispersive
521
spectra (EDS) of the spots 1 and 2 in the SEM micrographs.
522 523 524
525
ACCEPTED MANUSCRIPT
Table 1 Average length (mm)
Average width (µm)
Aspect Ratio
Fibrous material (106 fibers/g)
Content of fines by mass (%)
0.83 ± 0.05
16.4 ± 0.2
51
18.17 ± 1.11
25.7 ± 0.6
526 527
RI PT
528 529 530 531
SC
532 533 534
M AN U
535 536 537 538 539
543 544 545 546 547 548 549 550 551 552 553 554 555 556
EP
542
AC C
541
TE D
540
558 559
ACCEPTED MANUSCRIPT
Table 2 Raw material
Content (% by mass)
Portland cement (CPV-ARI) a
70
Limestone filler b
27
Unbleached eucalyptus pulp c
3
(a) ASTM C150 type I [19], provided by Caue, (b) Itaú, (c) Fibria S.A.
RI PT
557
560 561 562
SC
563 564 565
M AN U
566 567 568 569 570
574 575 576 577 578 579 580 581 582 583 584 585 586 587
EP
573
AC C
572
TE D
571
ACCEPTED MANUSCRIPT
588
590 591 592
Table 3 Structures
Elastic modulus (GPa)
Hardness (GPa)
Interfacial/transition zone
17.2 ± 5.6 c
0.4 ± 0.2 c
Bulk matrix
72.7 ± 14.3 b
5.3 ± 1.8 b
Anhydrous grain
89.9 ± 17.1 b
4.5 ± 1.9 b
Cellulosic fiber
144.3 ± 48.9 a
8.9 ± 2.5 a
1
Lowercase letters (a, b and c) in the same column represent comparisons between conditions of the composites. Different letters indicate significant differences amongst the properties of the composites.
593
SC
594 595
M AN U
596 597 598 599 600
605 606 607 608 609 610 611 612 613 614 615 616 617 618
EP
604
AC C
603
TE D
601 602
RI PT
589
ACCEPTED MANUSCRIPT
SC
RI PT
619
621
M AN U
620 Figure 1
622 623 624
628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
EP
627
AC C
626
TE D
625
ACCEPTED MANUSCRIPT
643
RI PT
Interface
SC
Bulk matrix
M AN U
644 645 646
Figure 2
647 648 649
653 654 655 656 657 658 659 660 661 662 663 664 665 666
EP
652
AC C
651
TE D
650
ACCEPTED MANUSCRIPT
RI PT
667
SC
5
µm 668 669 670
Figure 3
671 672
676 677 678 679 680 681 682 683 684 685 686 687 688 689
EP
675
AC C
674
TE D
673
M AN U
10
ACCEPTED MANUSCRIPT
690
Anhydrous grain
B
SC
Bulk matrix
RI PT
A
691 692 Figure 4
M AN U
693 694 695 696 697
701 702 703 704 705 706 707 708 709 710 711 712 713
EP
700
AC C
699
TE D
698
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
(a)
AC C
EP
TE D
(b)
714 715
(c) Figure 5
ACCEPTED MANUSCRIPT
Transition zone
81
Cellulosic fibre Anhydrous grain
SC
71
RI PT
101 91
Bulk matrix
(b)
AC C
EP
TE D
M AN U
(a)
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
(c) 716 717
Figure 6
718
722 723 724 725 726 727 728 729 730 731 732 733 734 735
EP
721
AC C
720
TE D
719
ACCEPTED MANUSCRIPT
736
1 Hydration products in the cellulosic fiber
1 Cellulosic fiber
SC
(a)
(b)
737 738
Anhydrous grain
M AN U
30 µm
Bulk matrix
Transition zone
RI PT
2
Figure 7
AC C
EP
TE D
739
2