1 Influence of type and dosage of micro-fibres on the physical properties of fibre reinforced mortar

2 matrixes

3 T. Simõesa,b, H. Costaa,c,*, D. Dias-da-Costad,e, E. Júlioa,b

a
4 CERIS, Instituto Superior Técnico, Universidade de Lisboa, Portugal.
b
5 Department of Civil Engineering, Architecture and Georesources, Instituto Superior Técnico, Universidade
6 de Lisboa, Portugal.
c
7 Department of Civil Engineering, Instituto Superior de Engenharia de Coimbra, Instituto Politécnico de
8 Coimbra, Portugal.
d
9 School of Civil Engineering, The University of Sydney, Australia.
e
10 ISISE, Departamento de Engenharia Civil, Universidade de Coimbra, Portugal.

11 *Corresponding author; e-mail address: [email protected]

12

13 Abstract

14 An experimental study is herein presented focusing the physical properties of fibre-reinforced mortar. A
15 reference mortar mixture was adopted and three different types of reinforcement with micro fibres (steel,
16 polypropylene and glass) were adopted, adding different volumetric dosages, ranging from 0.5% to 2%, in
17 0.5% increments. The influence of each type of fibre and dosage on the properties of the mixture, including
18 workability, density, porosity, and Young’s modulus was analysed. In summary, it was observed that the
19 workability was extremely reduced for polypropylene and glass fibres. It was found that this parameter has
20 an important role in terms of porosity and bulk density. The latter property decreased until 8% when
21 polypropylene or glass fibres were added to the reference mixture. However, the porosity was significantly
22 higher for mixtures with those fibres, with values reaching nearly twice the porosity of the reference mixture.
23 The dynamic Young’s modulus was not highly sensitive to the presence of steel fibres, with a reduction
24 lower than 5% in all tested dosages. In the case of mixtures reinforced with polypropylene fibres and glass
25 fibres, this reduction has reached at most 16% and 13%, respectively. This effect was even higher for the
26 static Young’s modulus, experiencing a maximum reduction of 32%, for a 0.5% dosage of polypropylene
27 fibres.

28

29 Keywords: Fibre reinforced mortar; steel, polypropylene, and glass fibres; Young’s modulus; workability;
30 porosity.

Simões, T., Costa, H., Dias-da-Costa, D., Júlio, E., Influence of type and dosage of micro-fibres on the
physical properties of fibre reinforced mortar matrixes, Construction and Building Materials, Elsevier, 187:
1277–85, 2018 (doi: 10.1016/j.conbuildmat.2018.08.058).
31 1. Introduction
32 In ancient civilizations, natural fibres were adopted to reinforce brittle construction materials, such as earth
33 bricks [1]. In modern times, the industrial use of fibres in construction only appeared after the development
34 of the Hatscheck machine in 1900, which allowed the continuous production of sheets using an asbestos-
35 cement composite [2]. During the 1960s and 1970s, the use of asbestos fibres decreased with the awareness
36 of the health problems caused by long-term heavy exposure to these airborne fibres [3]. Since then, the fibre
37 reinforced concrete technology has evolved and both new applications and new fibres have emerged.
38 Different materials, such as steel, polypropylene, and glass, among others, gradually widespread, in
39 particular for the production of structural fibre reinforced concrete (FRC) [2]. Several types of fibres and the
40 corresponding main properties are listed on Table 1 [4].

41 Table 1 – Typical properties of fibres

Diameter Specific Young's modulus Tensile strength Elongation at break
Fibre
(µm) density (GPa) (GPa) (%)
Steel 5–500 7.84 200 0.5–2.0 0.5–3.5
Glass 9–15 2.6 70–80 2–4 2.0–3.5
Crocidolite (asbesto) 0.02–0.40 3.4 196 3.5 2.0–3.0
Chrysolite (asbesto) 0.02–0.40 2.6 164 3.1 2.0–3.0
Basalt* 20 2.65 93–110 4100–4800 3.1–3.2
Polypropylene 20–400 0.90–0.95 3.5–10.0 0.45–0.76 15–25
Aramid (kevlar) 10–12 1.44 63–120 2.3–3.5 2.0–4.5
Carbon 8–9 1.6–1.7 170–600** 2.5–4.0 0.5–1.5
Nylon 23–400 1.14 4.1–5.2 0.75–1.00 16.0–20.0
Cellulose – 1.2 10 0.3–0.5 –
Acrylic 18 1.18 14.0–19.5 0.4–1.0 3
Polyethylene 25–1000 0.92–0.96 5 0.08–0.60 3–100
Wood fibre – 1.5 71 0.9 –
Sisal 10–50 1.5 – 0.8 3
Cement matrix (for comparison) – 1.5–2.5 10–45 0.003–0.007 0.02
*[5]
**[6]
42

43 The use of fibres in concrete mixtures has impact on both fresh-state and hardened state concrete properties,
44 namely on workability, strength, and ductility, but they are mainly used to enhance structural performance.
45 The most common fibres capable of improving the properties of plain concrete are made of steel, glass or
46 polypropylene. Most existing research has been focused on steel fibres. However, there are still contradictory
47 results concerning the prediction of the material properties. Studies seem to suggest that the compressive
48 strength of steel FRC [7,8] increases up to 10% when compared to plain concrete, whereas other studies
49 claim that this change is only marginal or even not related with the introduction of fibres [9,10].
50 Nevertheless, all studies unanimously refer to the enhanced toughness, ductility and flexural strength of the
51 steel FRC, the latter reaching increases ranging from 30 to 125% compared to plain concrete, depending of
52 concrete strength and dosage [7–12]. In addition, the failure mode of steel FRC changes from brittle to
53 ductile and the post-cracking response is significantly improved [8,9]. However, before reaching this regime,
54 the Young’s modulus does not seem to be significantly affected by the steel fibres [9,10].

55 Research on FRC other than steel reinforcement is limited. Nevertheless, in the case of polypropylene FRC,
56 some studies [13–16] report increased flexural toughness and ductility, compared to plain concrete, for both
57 lower and higher dosages, increasing with the percentage of reinforcement, being influenced by the
58 properties of both fibres and concrete. It should be highlighted that some studies [14,17,18] mention the
59 compressive strength of polypropylene FRC to be nearly unchanged by the fibres, whereas others [15,19]
60 show an increase up to 20%. In terms of flexural strength, some authors [14,17] report no impact on this
61 material property, whereas others [15,16,18] state an increase of 10% maximum, or even a decrease on this
62 property [13], in this case attributed to higher air content in the FRC.

63 Most studies focused on glass fibre reinforced concrete mention an increase in the compressive strength
64 ranging from 20 to 25% [20–24], although other publication [25] suggest a marginal decrease of this
65 parameter. Regarding the flexural strength, the majority of manuscripts [20–25] point out to an increase of
66 15 to 20%, together with an improved toughness of glass FRC, when compared to plain concrete. In what
67 concerns the Young’s modulus, results are contradictory [24,25].

68 If the Young’s modulus of the fibres is higher than that of the reference plain concrete, the resulting FRC
69 shall exhibit increased stiffness [26]. However, there are other influencing factors that can lead to a different
70 outcome. In fact, it is known that the addition of fibres can impact on the workability of the mixture [27,28]
71 and lead to an increase of the material porosity. Being the latter responsible for an important reduction of the
72 Young’s modulus, this can explain the allegedly contradictory conclusions found in the literature.

73 The design of concrete structures requires good predictive material models that ultimately should take into
74 account the presence of fibres within the mixture. In existing codes, concrete is frequently characterised by
75 the compressive strength, whereas other properties, such as the Young’s modulus are estimated using
76 correlations based on the first. According to [29], concrete porosity and Young’s modulus are related by the
77 equation En = (1 – p)2, where En is the normalized effective Young’s modulus and p the porosity. Concrete
78 density can be estimated from the porosity and properties of constituent materials according to equation μ =
79 (100 - fp) / 100 × μ m + fp / 100 ×μf, where µ is the specific density, fp is the fibre percentage, μ m and μf are the
80 values of specific density of the concrete matrix and of the fibres, respectively. Nevertheless, the great
81 majority of studies found in the literature [7,9,10,13,16,17,25] on FRC are essentially focused on a single
82 type of fibre and corresponding mechanical behaviour. In addition, different authors adopt different mixtures
83 and test set-ups, which makes it difficult to compare and generalize results and conclusions, particularly in
84 what concerns different types of fibres.

85 Concrete matrixes includes three phases (binding matrix, interfacial transition zones, and coarse aggregates).
86 There are several studies on concrete at both macro and micro-scale but very few at meso-scale, herein
87 understood as the interfacial transition zones - ITZ. Moreover, these studies on the ITZ are mainly focused
88 on chemical connections. The present work does not focus on a specific practical application, on contrary it
89 is part of a larger project which aims to properly understand the behaviour of both the aggregate/binding
90 matrix and the fibre/binding matrix interfaces in FRC, following a physical approach. Taking this into
91 consideration, this work aims to present an extensive comparative experimental study focused on the
92 reinforcement of a reference mortar matrix, using three different types of micro fibres (polypropylene, glass
93 and steel).
 94 The purpose was to assess how different fibres influence the physical properties of the FRC. The following
95 specific objectives were defined:

96 - to assess how the FRC density varies with the fibre type and dosage in the mixture;
97 - to characterise both static and dynamic Young’s modulus for the FRC and to assess how it changes
98 with the fibre type and dosage in the mixture;
99 - to determine the porosity for all mixtures and to identify how this parameter may influence the
100 results of the remaining studied properties;
101 - provide a basis of comparison that allows a choice of fibre type depending on the structure where the
102 FRC will be applied.

103 The paper is organised as follows: it starts with the introductory section 1; then, section 2 describes the
104 experimental programme, including all the mixtures and adopted specimens; in section 3, the main results are
105 presented and the physical properties of the material are discussed; lastly, section 4 lists the main conclusions
106 of this research.

107 2. Experimental Programme

108 The experimental programme was outlined taking into account the different aims set in the previous section.
109 In the following, the geometry and number of specimens, mixtures and test set-ups for the characterisation of
110 physical properties, are described.

111 2.1. Specimens and material properties

112 Four different mortar cases were adopted in this research work, being the first one composed of the reference
113 self-compacting cementitious plain matrix (without fibres), whereas the other three cases were based on
114 reinforcement micro fibres, of steel, polypropylene and glass. Four different dosages of steel fibres were
115 assumed, in proportions ranging from 0.5% up to 2%, in increments of 0.5%. The reference mixture was
116 intentionally designed as self-compacting in order to balance the expected loss of workability when
117 significant dosages of micro fibres are added to the matrix, in particular for those of glass and polypropylene.

118 That loss of workability is due to the number of fibres (per litre of fibres), which number is very high for
119 polypropylene, and even higher for glass, when compared to the number found when the same dosage of
120 steel fibres is adopted. This is explained by the reduced cross-sectional area of the formers compared to the
121 latter. In this regard, Figure 1 shows the number of fibres, per litre of fibre material. Taking this into account,
122 one can easily understand that the increased number of fibres can significantly reduce the workability of the
123 mixture for higher dosages of polypropylene and glass. This is the reason why the maximum dosages found
124 in the literature [27,28] regarding these fibres are typically lower than those regarding steel fibres.
125 Nevertheless, and for comparison purposes, the experimental programme herein described was carried out
126 using the same volumetric proportions for all types of fibres, i.e., ranging from 0.5% to 2%. The expected
127 loss of self-compacting and flowability in mixtures reinforced with polypropylene and glass micro fibres was
128 predicted in advance. That loss was further experimentally proved and the necessary vibration was
129 performed, using the compacting table, until flowing and compacting processes were complete, and no
130 additional air voids were released from the mixtures in the moulds, assuring that the coherency of the matrix
131 parameters was maintained and the comparative analysis is possible and valid between mixtures.

Number of fibres (108)

4

0
Steel Polypropylene Glass
Type of fibre
132

133 Figure 1 – Number of fibres per litre

134

135 For each mixture, twenty-four prismatic specimens were produced according to EN 196 [30], allowing a
136 statistical study to be performed, leading to a total of three hundred and twelve samples, defined in Table 2.

137 Table 2 – Number of specimens produced for each case

Dosage (%)
Type of fibre
0.0 0.5 1.0 1.5 2.0
None 24 - - - -
Steel - 24 24 24 24
Polypropylene - 24 24 24 24
Glass - 24 24 24 24
Total specimens 312
138

139 Considering the high number of specimens considered in this study, and since coarse aggregates were not
140 included, instead of using the geometry proposed by RILEM TC162-TDF [31], it was decided to adopt
141 smaller specimens, 40×40×160 mm3 prisms. In addition, two 100×100×400 mm3 specimens were produced
142 for each mixture with the purpose of characterising the static Young’s modulus according to [32].

143 2.2. Mixtures and production of specimens

144 The reference binding matrix, common to all the mixtures, was produced using cement CEM II/A-L 42.5R,
145 limestone filler, third generation superplasticiser (eter-polycarboxylates based), and water. Due to the
146 geometry of the specimen, the maximum aggregate size was limited to siliceous medium oven-dried sand
147 (0/4 mm). As mentioned before, the fibres were made of steel (Dramix® OL 13/.20), polypropylene
148 (Vimafibre 512), or glass (Vimacrack). The corresponding properties are listed in Table 3.

149 Table 3 – Main properties for all the fibres adopted

Diameter Length Specific Young's modulus Tensile strength
Type of fibre
(µm) (mm) density (GPa) (MPa)
Dramix® OL 13/.20 200 13 7.84 200 2600
Vimafibre 512 34–45 12 0.91 3.5–4.0 340–400
Vimacrack 14 12 2.68 72 1700
150

151 The reference mixture was designed using the procedure described in [33], based on the Feret’s expression to
152 predict the strength of the binding paste. Both the Feret’s coefficient and the air content were first
153 determined in a preliminary test mixture (Figure 2). Afterwards the mixture was progressively enhanced until
154 the final formulation was set, for a target compressive strength of the reference mixture of 65MPa.

(a) (b)
Figure 2 – Preliminary stage: a) mixer; and b) aerometer.
155

156 Table 4 shows the final composition for each case.

157 Table 4 – Final mixtures proportion (kg per cubic meter)

Constituents
Mixtures CEM II/A-L Limestone BASF Medium Fibres*
Water Glenium
42.5R filer Sand St Po Gl
Sky 526
Reference 1340.6 - - -
St 0.5 1327.7 39.3
St 1.0 1314.8 78.5
- -
St 1.5 1301.1 117.8
St 2.0 1288.2 157.0
Po 0.5 1327.7 4.6
Po 1.0 475.0 285.0 197.6 6.2 1314.8 9.1
- -
Po 1.5 1301.1 13.7
Po 2.0 1288.2 18.2
Gl 0.5 1327.7 13.4
Gl 1.0 1314.8 26.8
- -
Gl 1.5 1301.1 40.2
Gl 2.0 1288.2 53.6
* ‘St’ stands for ‘Steel’, ‘Po’ stands for ‘Polypropylene’ and ‘Gl’ stands for glass.
158

159 The reinforced mixtures were produced by first mixing the constituents of the reference mixture (cement,
160 filler, sand, water, superplasticiser), then gradually adding the fibres, and finally mixing everything together
161 until reaching an homogeneous texture (Figure 3) [2]. Next, the specimens were cast and the formwork was
162 removed after 24 hours. Lastly, the specimens were cured in water at 20 ± 2°C [30], and removed and dryed
163 approximately 24 hours before testing.
 (a) (b)
Figure 3 – Mixing procedure: a) before; and b) after adding the fibres.
164

165 Six additional specimens were produced with the reference mixture for assessing the mortar compressive
166 strength at 28 days (Figure 4a). An average value of 67.7MPa and a standard deviation of 2.8 MPa was
167 found.

(a) (b)
Figure 4 – a) compressive tests of the reference mixture specimens;
b) helium pycnometer.

168

169 2.3. Characterisation of the physical properties

170 The bulk density was calculated by dividing the mass of each specimen, obtained with a weighing-machine,
171 by its known volume. The specific density was measured using a helium pycnometer, shown in Figure 4b
172 [34], and the used powder samples, of about 2,5 cm3, were randomly extracted from the specimens. The
173 porosity of the mixtures was calculated, according to the following equation:

𝜇−𝜋
𝑝= (1)
𝜇

174 where 𝑝 is the porosity, 𝜇 and 𝜋 are respectively the specific and bulk densities.

175 The dynamic Young’s modulus was determined using two different methods (Figure 5a and 5b): The
176 ‘Pundit’, based on the speed of propagation of ultrasound waves through a solid [35], which was used in all
177 specimens; and a method based on the resonance frequency of the solid [36], which was applied to three
178 randomly selected specimens of each mixture.

(a) (b) (c)

Figure 5 – Measurement of the Young’s modulus using: a) Pundit; b) resonance; and c) static.
179

180 The static Young’s modulus was determined directly from the ratio between the stress applied to the
181 prismatic specimen and the corresponding deformation (Figure 5c) as defined in the standard [32].

182 3. Results and discussion

183 In the following sub-sections, results addressing workability, density, porosity, and Young’s modulus are
184 presented and discussed.

185 3.1. Workability, density and porosity

186 During mixing, the reduction of workability was clear for polypropylene and glass fibres mixtures. Flow
187 table tests [37] were performed on the mixtures, where the reference and the steel fibres reinforced mixtures
188 presented self-compacting flowability, whereas the mixtures with polypropylene and glass fibres presented
189 well moulded plastic consistency on the test specimens, with no significant flow values, thus showing
190 reduced workability. Considering that difficulty of measure the workability by normalised methods, this
191 property was then qualitatively accessed. This reduction can be attributed to the high number of fibres in
192 those mixtures, creating a dense skeleton and affecting the rheology and the internal flow movements of the
193 fresh state mortar matrix.

194 The specific and bulk densities and porosity, as the confidence interval (CI) at 95% are listed in Figure 6a
195 and plotted in Figures 6b-d. For each type of fibre, the specific density changed proportionally with the
196 dosage (see Fig. 6b). Comparing with the reference mixture that has a specific density of 2.60, one sees this
197 parameter increased in the presence of steel fibres, ranging from 2.57 to 2.65 (depending on the dosage), and
198 decreased for polypropylene fibres, ranging from 2.54 to 2.49 (for 1.0 and 2.0% respectively). In the case of
199 glass fibres, this parameter slightly increases.

200 Comparing to the reference mixture, the bulk density decreased when polypropylene or glass fibres were
201 adopted. In fact, the reference density was 2.38, whereas the minimum and maximum densities for
202 polypropylene were, respectively, 2.18 and 2.30, and the corresponding values for the glass fibres were 2.26
203 and 2.31. This represents a maximum reduction of approximately 5%, in the case of glass fibres with a
204 dosage of 2%, and 8% in the case of polypropylene fibres with a dosage of 0.5%.

205 For the mixtures based on steel fibres, the bulk density increased proportionally with the dosage up to 1.5%
206 (from 2.35 to 2.42). For the highest dosage (2%) the bulk density decreased due to the extremely reduced
207 workability found in this mixture. However, this parameter does not change significantly, only varying from
208 -1.3% to 1.8%

209 The porosity of the reinforced mixtures is often higher than the one measured on the reference plain mixture
210 (8.5%). This effect is more clear in both polypropylene and glass fibres, where the porosity ranges, with
211 increasing dosage, from 13.0% to 8.0%, in the case of polypropylene fibres, and from 8.9% to 11.5%, in the
212 case of glass fibres. For glass fibres, the porosity evolution follows the expected trend but, for polypropylene
213 fibres, a different behavior is observed. A possible explanation for the porosity decrease with the increase of
214 fibre content, in the case mixtures with polypropylene fibres, is the high flexibility of those fibres. It is
215 possible that, during mixing and casting, some fibres easily curl filling in some voids, which combined with
216 the additional required vibration to assure compacting, results on porosity reduction. This process tends to
217 increase with the dosage of fibres as there are more available fibres to fill in the voids. However, for the
218 mixtures with glass fibre that behaviour does not occur, because those are stiffer and the higher number of
219 fibres creates an increased dense skeleton where vibration is less effective, with the consequence of porosity
220 increase. For steel fibre this parameter slightly decreases up to an adopted dosage of 1.5% and then increases
221 again up to 2.0%. The correct explanation for that is the high stiffness of steel fibres, combined with the
222 matrix viscosity, which greatly increase the shear mixing forces, resulting on improved flowability and high
223 capacity of air release of the matrix, reducing the porosity. However, this is effective until a certain fibres
224 dosage (in this case around 1,5%), reversing the trend after that, above which dosage the fibres mesh
225 becomes so tangled that increases the voids in the matrix.

226

227

228

229
 Specific CI Bulk CI Porosity 3,0
Mixture
Density (95%) Density (95%) (%) 0.5% 1.0% 1.5% 2.0% Ref

Reference 2.603 0.001 2.380 0.009 8.54

St 0.5 2.571 0.003 2.349 0.013 8.62
St 1.0 2.586 0.010 2.367 0.009 8.46 2,0

Specific density
St 1.5 2.621 0.006 2.423 0.011 7.56
St 2.0 2.654 0.014 2.399 0.019 9.61
Po 0.5 2.510 0.010 2.184 0.019 12.96
Po 1.0 2.537 0.003 2.233 0.014 11.96
1,0
Po 1.5 2.526 0.006 2.300 0.012 8.95
Po 2.0 2.490 0.007 2.291 0.014 7.98
Gl 0.5 2.534 0.005 2.309 0.012 8.87
Gl 1.0 2.531 0.010 2.283 0.010 9.80
Gl 1.5 2.535 0.011 2.286 0.011 9.82 0,0
Steel Polypropylene Glass
Gl 2.0 2.553 0.011 2.259 0.013 11.52

(a) (b)

3.0 20
0.5% 1.0% 1.5% 2.0% Ref 0.5% 1.0% 1.5% 2.0% Ref

16

2.0

Porosity (%)
Bulk density

12

8
1.0

0.0 0
Steel Polypropylene Glass Steel Polypropylene Glass

(c) (d)
Figure 6 – (a) Table containing the average densities and porosity for all mixtures; and corresponding charts for (b)
specific density; (c) bulk density; and (d) porosity.
230

231 3.2. Young’s modulus and compressive strength

232 The results concerning the dynamic and static Young’s modulus, as the confidence interval (CI) at 95% are
233 listed in Figure 7a and plotted in Figures 7b-d. As a general observation, the dynamic results obtained using
234 the resonance method present higher variability than the ones obtained using the Pundit method. However,
235 the average values from both methods are quite similar (as one can check by comparing e.g. Figs 7b and c).
236 It should be noted that the trends are very similar in both methods. However, the values from Resonance
237 method are always lower than those from Pundit method. Although these two methods evaluate the dynamic
238 Young’s modulus, the physical phenomenon associated to each one is different. While the Pundit method is
239 based on the propagation speed of an ultrasonic wave within a solid, the Resonance method is based on the
240 vibration frequency of a solid. In the calculations used in the Resonance method the Poisson’s ratio has to be
241 admitted and in this study a value of 0.2 was chosen. If the real Poisson’s ratio is slightly different, the result
242 will be affected and this explains the difference in the methods. However in cementitious materials the
243 Pundit method is the most commonly used and the Resonance method was used to compare with the first.
244 The dynamic Young’s modulus was always higher than the corresponding static measurement, in accordance
245 with [38–40].
246 For the reference mixture, the static and dynamic Young’s modulus (obtained with the Pundit method) are
247 40.77 and 50.31 GPa, respectively. Compared with the reference mixture, nearly all reinforced mixtures
248 show a lower Young’s modulus.

249 For steel fibres, the dynamic Young’s modulus (obtained with the Pundit method) is close to the value
250 measured in the reference mixture, except for the highest dosage where a reduction of 5% (relatively to the
251 reference mixture) is observed. This result is in agreement with results reported in other works [9,10], where
252 the modulus of elasticity does not seem to change significantly with the presence of steel fibres.

253 In the case of polypropylene fibre, the reduction is much higher, reaching nearly 16% for the lower dosage.
254 For other dosages, the dynamic modulus increases slightly up to 1.5% dosage, in which case the reduction
255 changes to 11%. In the case of glass fibre, this parameter decreases with the dosage of fibres, from 47.15 to
256 43.60 GPa, which represents a reduction of 6% and 13% relatively to the reference mixture.

257 Similar conclusions can be drawn for the static Young’s modulus, although in this case the reduction is
258 significantly higher. For example, comparing to the reference mixture, there is a reduction on average of
259 2.5%, 13% and 9.5% for steel, polypropylene and glass fibres in the dynamic Young’s modulus, whereas the
260 corresponding static modulus decreases on average 8.5, 17 and 14%, respectively. This parameter seems to
261 be linked to the porosity. In this case the evolution is in the opposite direction of the porosity, as expected.
262 The fibre orientation may have an effect on the experimental readings because the static module is measured
263 in one direction of the test specimens and the quantity of fibres orientated on that direction can influence the
264 result. The obtained values suggest that the fibres dispersion is acceptable.

265 The compressive and flexural strengths, determined by [30], are shown in Figure 8a, where fcm is the mortar
266 average compressive strength. These parameters were used to correlate the compressive strength and the
267 dynamic Young’s modulus, represented in Figure 8b. It can be seen that a linear trend line can be used to
268 predict the dynamic Young’s modulus from the compressive strength with an acceptable degree of precision
269 (R2 = 0.84). The compressive strength tends to increase with the increase of fibres addition. For
270 polypropylene fibres, the strength found with the 0.5% fibre content is 10% lower than the reference, whilst
271 all other reinforcement ratios show higher strengths. For glass fibres, the strength was always higher
272 comparing to the reference. However, the steel fibres were those that most influence the compressive
273 strength, with an increase of approximately 30% for the 1.5% of reinforcement.
 Ed CI Ed CI Ee CI 60
Mixture 0.5% 1.0% 1.5% 2.0% Ref
Pundit (95%) Resonance (95%) (GPa) (95%)

Dynamic Young's modulus (GPa)

Reference (GPa)
50.31 0.33 (GPa)
48.45 0.31 40.60 0.35 50
St 0.5 48.98 0.52 47.99 1.76 35.76 0.54
St 1.0 49.04 0.32 46.71 1.37 38.58 0.31 40
St 1.5 50.33 0.35 47.46 1.71 40.60 0.35
St 2.0 47.94 0.78 45.82 1.61 33.44 0.47 30
Po 0.5 42.22 0.78 39.98 0.82 27.67 0.48
Po 1.0 43.47 0.37 43.11 0.99 34.95 0.26 20
Po 1.5 44.79 0.32 43.87 1.52 35.48 0.27
Po 2.0 44.01 0.34 43.61 0.50 36.89 0.29
10
Gl 0.5 47.15 0.28 45.32 1.23 36.89 0.58
Gl 1.0 46.67 0.27 44.21 1.02 35.48 0.31
Gl 1.5 44.58 0.33 42.48 0.83 34.95 0.26 0
Steel Polypropylene Glass
Gl 2.0 43.60 0.30 41.18 1.52 32.51 0.22
(a) (b)
60 60
0.5% 1.0% 1.5% 2.0% Ref 0.5% 1.0% 1.5% 2.0% Ref
Dynamic Young's modulus (GPa)

50 50

Static Young's modulus (GPa)

40 40

30 30

20 20

10 10

0 0
Steel Polypropylene Glass Steel Polypropylene Glass

(c) (d)
Figure 7 – (a) Table containing the average dynamic and static Young’s modulus for all mixtures; and corresponding
charts for (b) dynamic modulus (pundit); (c) dynamic modulus (resonance); and (d) static modulus.
274

Compressive Flexural 52
Mixture y = 0.26 × fcm + 26.24
Strength Strength
Dynamic Young's modulus (GPa)

R² = 0.84
50
(MPa) (kN)
Reference 67.72 5.25 48
St 0.5 87.58 5.89
St 1.0 88.53 9.20 46
St 1.5 91.96 10.96
St 2.0 84.11 13.71 44
Po 0.5 60.16 3.41 42
Po 1.0 68.51 3.60
Po 1.5 72.46 3.80 40
Po 2.0 73.67 4.54
38
Gl 0.5 70.92 4.25 50 60 70 80 90 100
Gl 1.0 78.40 4.62 Compressive strength (MPa)
Gl 1.5 75.84 5.23
Gl 2.0 72.79 5.71

275 Figure 8 – (a) Table containing the average compressive and flexural strengths for all mixtures; and (b) dynamic
276 Young’s modulus vs. compressive strength

277 4. Conclusions
278 This research aimed at studying the influence of steel, polypropylene, and glass micro fibres on the physical
279 properties of FRC. The cementitious matrix was kept constant to facilitate comparison of different material
280 properties, namely workability, density, porosity, and Young’s modulus, only varying the type and dosage of
281 fibres, the latter ranging from 0.5% up to 2%, in 0.5% increments. A total of three hundred and twelve
282 samples, 24 for each mixture, were produced and tested.

283 As a general conclusion, it can be stated that the workability of the mixture is severely reduced by the fibres,
284 particularly for mixtures with polypropylene and with glass fibres, since the number of fibres in these cases
285 is, for the same dosage, much higher than the corresponding value for mixtures with steel fibres. Specific
286 conclusions are highlighted next:

287 – The steel fibres have higher density than the cementitious matrix, thus the specific density of the mortar
288 increases with its dosage increase. For the polypropylene fibres the tendency is the opposite, decreasing with
289 the introduction of fibres, since the density of fibres is lower than that of the cementitious matrix. In the case
290 of glass fibres, this parameter exhibits marginal changes, since the density of the fibres and that of the
291 cementitious matrix are very similar. The bulk density of the reinforced mixes is more influenced by the
292 porosity than by the reinforcement density. This is clear for the mixture of polypropylene fibre with 0.5% of
293 reinforcement, for which the reduction on the specific density is only 0.3%, while the bulk density decreases
294 about 8% (see Figure 6c). The increased porosity has much higher impact on the bulk density – particularly
295 for polypropylene and glass fibres – than the theoretical relation between bulk density and fibre/matrix
296 density ratio.

297 – The porosity of the reinforced mixtures is often higher than the one measured on the reference plain
298 mixture. It can be concluded from Figure 1 that the modification of the type of fibre implies a significant
299 change in the number of fibres within the mixture. Therefore, the workability is highly affected and the
300 compacting capacity of reinforced mixtures is greatly reduced. Consequently, many air pockets that would
301 have been released in normal circumstances, remain trapped in the FRC matrix, leading to the generation of
302 air voids in hardened state and thus to an increased porosity.

303 – The dynamic and static Young’s moduli for reinforced mixtures are lower in comparison with the reference
304 mixture. The higher porosity (Figure 6d) in these mixtures seems to have a determinant role, promoting
305 enormous influence on the reduction of the Young’s modulus of the FRC matrix, in comparison to the
306 moduli of the fibres. The fibre/matrix interface can also have influence on results, because mortar interfaces
307 tend to have different properties from the rest of the mortar matrix.

308 – The compressive strength for polypropylene and steel fibres seems to be strongly linked to porosity,
309 showing an increase while the latter decrease. For glass fibres, the compressive strength only increases up to
310 the ratio of 1%, decreasing proportionally to the fibre content after this dosage. This singularity most
311 probably occurs because of the loss on workability and of the air increase in the concrete matrix.

312 Acknowledgements
313 Authors would like to acknowledge the Portuguese Foundation for Science and Technology (FCT) by
314 funding the project PTDC/ECM/119214/2010, entitled Concrete Behaviour at Meso-Scale. The first author
315 acknowledges the support of FCT through the individual grant SFRH/BD/84355/2012, and the third author
316 acknowledges the support of the Australian Research Council through the Discovery Early Career
317 Researcher Award (DE150101703) and of the Faculty of Engineering & Information Technologies of the
318 University of Sydney, through the Faculty Research Cluster Program. Lastly, authors acknowledge the
319 support of the following companies that offered consumable items used in the experimental study, namely
320 Secil, Argilis, BASF, Omya, Vimaplás and Bekaert.

321 References
322 [1] M. Vitruvius Pollio, D. Barbaro, I dieci libri dell’architettura, Vinegia, 1556. doi:10.3931/e-rara-7582.

323 [2] A.C.I. 544.1R-96, State-of-the-Art Report on Fiber Reinfirced Concrete, 2002.

324 [3] R. Virta, Asbestos: Geology, Mineralogy, Mining, and Uses, U.S. Geological Survey, 2003.

325 [4] A. Bentur, S. Mindess, Fibre Reinforced Cementitious Composites, Taylor & Francis, 2007.

326 [5] C. Jiang, K. Fan, F. Wu, D. Chen, Experimental study on the mechanical properties and microstructure of
327 chopped basalt fibre reinforced concrete, Mater. Des. 58 (2014) 187–193.
328 doi:http://dx.doi.org/10.1016/j.matdes.2014.01.056.

329 [6] K.K. Chawla, Composite Materials: Science and Engineering, Springer Science & Business Media, 2013.

330 [7] P.. Song, S. Hwang, Mechanical properties of high-strength steel fiber-reinforced concrete, Constr. Build.
331 Mater. 18 (2004) 669–673. doi:10.1016/j.conbuildmat.2004.04.027.

332 [8] J. Thomas, A. Ramaswamy, Mechanical Properties of Steel Fiber-Reinforced Concrete., J. Mater. Civ. Eng. 19
333 (2007) 385–392.

334 [9] R.S. Olivito, F.A. Zuccarello, An experimental study on the tensile strength of steel fiber reinforced concrete,
335 Compos. Part B Eng. 41 (2010) 246–255. doi:10.1016/j.compositesb.2009.12.003.

336 [10] Y. Şahin, F. Köksal, The influences of matrix and steel fibre tensile strengths on the fracture energy of high-
337 strength concrete, Constr. Build. Mater. 25 (2011) 1801–1806. doi:10.1016/j.conbuildmat.2010.11.084.

338 [11] N. Buratti, C. Mazzotti, M. Savoia, Post-cracking behaviour of steel and macro-synthetic fibre-reinforced
339 concretes, Constr. Build. Mater. 25 (2011) 2713–2722.

340 [12] N. Banthia, M. Sappakittipakorn, Toughness enhancement in steel fiber reinforced concrete through fiber
341 hybridization, Cem. Concr. Res. 37 (2007) 1366–1372.

342 [13] Y. Wu, Flexural strength and behavior of polypropylene fiber reinforced concrete beams, J. Wuhan Univ.
343 Technol. Sci. Ed. 17 (2002) 54–57. doi:10.1007/BF02832623.

344 [14] L. Bei-xing, C. Ming-xiang, C. Fang, L. Lu-ping, The mechanical properties of polypropylene fiber reinforced
345 concrete, J. Wuhan Univ. Technol. Sci. Ed. 19 (2004) 68–71. doi:10.1007/BF02835065.

346 [15] M. Hsie, C. Tu, P.S. Song, Mechanical properties of polypropylene hybrid fiber-reinforced concrete, Mater. Sci.
347 Eng. A. 494 (2008) 153–157. doi:10.1016/j.msea.2008.05.037.

348 [16] H. Cifuentes, F. García, O. Maeso, F. Medina, Influence of the properties of polypropylene fibres on the
349 fracture behaviour of low-, normal- and high-strength FRC, Constr. Build. Mater. 45 (2013) 130–137.
350 doi:10.1016/j.conbuildmat.2013.03.098.

351 [17] A.M. Alhozaimy, P. Soroushiad, F. Mirza, Mechanical Properties of Polypropylene Fiber Reinforced Concrete
352 and the Effects of Pozzolanic Materials, Cem. Concr. Compos. 18 (1996).

353 [18] H. Mazaheripour, S. Ghanbarpour, S.H. Mirmoradi, I. Hosseinpour, The effect of polypropylene fibers on the
354 properties of fresh and hardened lightweight self-compacting concrete, Constr. Build. Mater. 25 (2011) 351–
355 358. doi:10.1016/j.conbuildmat.2010.06.018.

356 [19] S. Kakooei, H.M. Akil, M. Jamshidi, J. Rouhi, The effects of polypropylene fibers on the properties of
357 reinforced concrete structures, Constr. Build. Mater. 27 (2012) 73–77. doi:10.1016/j.conbuildmat.2011.08.015.

358 [20] Z.O. Pehlivanlı, İ. Uzun, İ. Demir, Mechanical and microstructural features of autoclaved aerated concrete
359 reinforced with autoclaved polypropylene, carbon, basalt and glass fiber, Constr. Build. Mater. 96 (2015) 428–
360 433. doi:10.1016/j.conbuildmat.2015.08.104.

361 [21] i K. Chandramoul, P. Srinivasa Rao, N. Pannirselvam, T. Seshadri Sekhar, P. Sravana, STRENGTH
362 PROPERTIES OF GLASS FIBRE CONCRETE, J. Eng. Appl. Sci. 5 (2010).

363 [22] L.A. Qureshi, A. Ahmed, An Investigation On Strength Properties Of Glass Fiber Reinforced Concrete, Int. J.
364 Eng. Res. Technol. 2 (2013).

365 [23] C.S. Ravikumar, T.S. Thandavamoorthy, Glass Fibre Concrete: Investigation on Strength and Fire Resistant
366 Properties, J. Mech. Civ. Eng. 9 (2013).

367 [24] M.M. Abdullah, E.K. Jallo, Mechanical Properties of Glass Fiber Reinforced Concrete, AL Rafdain Eng. J. 20
368 (2012) 128–135.

369 [25] S.T. Tassew, A.S. Lubell, Mechanical properties of glass fiber reinforced ceramic concrete, Constr. Build.
370 Mater. 51 (2014) 215–224. doi:10.1016/j.conbuildmat.2013.10.046.

371 [26] S. Kurugöl, L. Tanaçan, H.Y. Ersoy, Young’s modulus of fiber-reinforced and polymer-modified lightweight
372 concrete composites, Constr. Build. Mater. 22 (2008) 1019–1028.
373 doi:http://dx.doi.org/10.1016/j.conbuildmat.2007.03.017.

374 [27] M. Sahmaran, A. Yurtseven, I. Ozgur Yaman, Workability of hybrid fiber reinforced self-compacting concrete,
375 Build. Environ. 40 (2005) 1672–1677. doi:http://dx.doi.org/10.1016/j.buildenv.2004.12.014.

376 [28] Y. Ding, S. Liu, Y. Zhang, A. Thomas, The investigation on the workability of fibre cocktail reinforced self-
377 compacting high performance concrete, Constr. Build. Mater. 22 (2008) 1462–1470.
378 doi:10.1016/j.conbuildmat.2007.03.034.

379 [29] K. Miled, Effective elastic properties of porous materials: Homogenization schemes vs experimental data,
380 Mech. Res. Commun. 38 (2011) 131–135.
381 [30] EN 196-1:1996. Methods of testing cement. Determination of strength., (1996).

382 [31] RILEM TC162-TDF: Test and Design Methods for Steel Fibre Reinforced Concrete. Bending test. Final
383 Recommendation., Mater. Struct. 35 (2002) 579–582.

384 [32] E 397-1993. Betões: Determinação do módulo de elasticidade em compressão, (1993).

385 [33] J. Lourenço, E. Júlio, P. Maranha, Betões de Agregados Leves de Argila Expandida, APEB, 2004.

386 [34] M. Viana, P. Jouannin, C. Pontier, D. Chulia, About pycnometric density measurements, Talanta. 57 (2002)
387 583–593. doi:http://dx.doi.org/10.1016/S0039-9140(02)00058-9.

388 [35] L. Qixian, J.H. Bungey, Using compression wave ultrasonic transducers to measure the velocity of surface
389 waves and hence determine dynamic modulus of elasticity for concrete, Constr. Build. Mater. 10 (1996) 237–
390 242. doi:http://dx.doi.org/10.1016/0950-0618(96)00003-7.

391 [36] ASTM, C 1259-94. Standard test method for dynamic Young’s modulus, shear modulus, and Poisson's ratio for
392 advanced ceramics by impulse excitation of vibration, (1994).

393 [37] EN 1015-3:1999. Methods of test for mortar for masonry. Determination of consistence of fresh mortar (by flow
394 table), (1999).

395 [38] Y. Zhou, J. Gao, Z. Sun, W. Qu, A fundamental study on compressive strength, static and dynamic elastic
396 moduli of young concrete, Constr. Build. Mater. 98 (2015) 137–145. doi:10.1016/j.conbuildmat.2015.08.110.

397 [39] K. Jurowski, S. Grzeszczyk, The Influence of Concrete Composition on Young’s Modulus, Procedia Eng. 108
398 (2015) 584–591. doi:10.1016/j.proeng.2015.06.181.

399 [40] BS 8110-2, Structural use of concrete-Part 2: Code of practice for special circumstances, Br. Stand. Inst. (1985).

400
401 List of Figures
402

Number of fibres (108)

4

0
Steel Polypropylene Glass
Type of fibre
403

404 Figure 1 – Number of fibres per litre

405

406

407

408

409

410

411

412
 (a) (b)
Figure 2 – Preliminary stage: a) mixer; and b) aerometer.
413

414

415

416

417

418

419

420

421

422
 (a) (b)
Figure 3 – Mixing procedure: a) before; and b) after adding the fibres.
423

424

425

426

427

428

429

430

431

432
 (a) (b)
Figure 4 – a) compressive tests of the reference mixture specimens; b) helium pycnometer.
433

434

435

436

437

438

439

440

441

442
 (a) (b)

(c)
Figure 5 – Measurement of the Young’s modulus using: a) Pundit; b) resonance; and c) static.
443

444

445

446

447

448

449

450

451
 Specific CI Bulk CI Porosity 3,0
Mixture
Density (95%) Density (95%) (%) 0.5% 1.0% 1.5% 2.0% Ref

Reference 2.603 0.001 2.380 0.009 8.54

St 0.5 2.571 0.003 2.349 0.013 8.62
St 1.0 2.586 0.010 2.367 0.009 8.46 2,0

Specific density
St 1.5 2.621 0.006 2.423 0.011 7.56
St 2.0 2.654 0.014 2.399 0.019 9.61
Po 0.5 2.510 0.010 2.184 0.019 12.96
Po 1.0 2.537 0.003 2.233 0.014 11.96
1,0
Po 1.5 2.526 0.006 2.300 0.012 8.95
Po 2.0 2.490 0.007 2.291 0.014 7.98
Gl 0.5 2.534 0.005 2.309 0.012 8.87
Gl 1.0 2.531 0.010 2.283 0.010 9.80
Gl 1.5 2.535 0.011 2.286 0.011 9.82 0,0
Steel Polypropylene Glass
Gl 2.0 2.553 0.011 2.259 0.013 11.52

(a) (b)

3.0 20
0.5% 1.0% 1.5% 2.0% Ref 0.5% 1.0% 1.5% 2.0% Ref

16

2.0

Porosity (%)
Bulk density

12

8
1.0

0.0 0
Steel Polypropylene Glass Steel Polypropylene Glass

(c) (d)
Figure 6 – (a) Table containing the average densities and porosity for all mixtures; and corresponding charts for (b)
specific density; (c) bulk density; and (d) porosity.
452

453

454

455

456

457

458
 Ed CI Ed CI Ee CI 60
Mixture 0.5% 1.0% 1.5% 2.0% Ref
Pundit (95%) Resonance (95%) (GPa) (95%)

Dynamic Young's modulus (GPa)

Reference (GPa)
50.31 0.33 (GPa)
48.45 0.31 40.60 0.35 50
St 0.5 48.98 0.52 47.99 1.76 35.76 0.54
St 1.0 49.04 0.32 46.71 1.37 38.58 0.31 40
St 1.5 50.33 0.35 47.46 1.71 40.60 0.35
St 2.0 47.94 0.78 45.82 1.61 33.44 0.47 30
Po 0.5 42.22 0.78 39.98 0.82 27.67 0.48
Po 1.0 43.47 0.37 43.11 0.99 34.95 0.26 20
Po 1.5 44.79 0.32 43.87 1.52 35.48 0.27
Po 2.0 44.01 0.34 43.61 0.50 36.89 0.29
10
Gl 0.5 47.15 0.28 45.32 1.23 36.89 0.58
Gl 1.0 46.67 0.27 44.21 1.02 35.48 0.31
Gl 1.5 44.58 0.33 42.48 0.83 34.95 0.26 0
Steel Polypropylene Glass
Gl 2.0 43.60 0.30 41.18 1.52 32.51 0.22
(a) (b)
60 60
0.5% 1.0% 1.5% 2.0% Ref 0.5% 1.0% 1.5% 2.0% Ref
Dynamic Young's modulus (GPa)

50 50

Static Young's modulus (GPa)

40 40

30 30

20 20

10 10

0 0
Steel Polypropylene Glass Steel Polypropylene Glass

(c) (d)
Figure 7 – (a) Table containing the average dynamic and static Young’s modulus for all mixtures; and corresponding
charts for (b) dynamic modulus (pundit); (c) dynamic modulus (resonance); and (d) static modulus.
459

460

461

462

463

464
 Compressive Flexural 52
Mixture Strength Strength y = 0.26 × fcm + 26.24

Dynamic Young's modulus (GPa)

R² = 0.84
(MPa) (kN) 50
Reference 67.72 5.25
St 0.5 87.58 5.89 48
St 1.0 88.53 9.20
St 1.5 91.96 10.96 46
St 2.0 84.11 13.71
Po 0.5 60.16 3.41 44
Po 1.0 68.51 3.60
Po 1.5 72.46 3.80 42
Po 2.0 73.67 4.54
40
Gl 0.5 70.92 4.25
Gl 1.0 78.40 4.62 38
Gl 1.5 75.84 5.23 50 60 70 80 90 100
Gl 2.0 72.79 5.71 Compressive strength (MPa)

465 Figure 8 – (a) Table containing the average compressive and flexural strengths for all mixtures; and (b) dynamic
466 Young’s modulus vs. compressive strength

467