Construction and Building Materials 124 (2016) 878–886

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Mechanical behaviour of basalt fibre reinforced concrete

John Branston a, Sreekanta Das a,⇑, Sara Y. Kenno b, Craig Taylor b
a
University of Windsor, Windsor, Ontario, Canada
b
MEDA Limited, Windsor, Ontario, Canada

h i g h l i g h t s

 Mechanical behaviour of BFRC was evaluated by flexural and impact testing.

 BFRC was made of (1) basalt bundled fibres and (2) basalt minibars.
 Both BFRC increased pre-cracking strength; had little effect under impact loading.
 Minibars significantly enhanced post-cracking behaviour but bundled fibres did not.
 Bundled fibres failed by rupturing, whereas the minibars failed by pulling out.

a r t i c l e i n f o a b s t r a c t

Article history: Chopped basalt fibre has recently gained popularity in concrete reinforcing applications due to its
Received 19 September 2015 environmentally friendly manufacturing process and excellent mechanical properties. The aim of this
Received in revised form 2 August 2016 research is to evaluate the relative merit of two types of basalt fibre (bundle dispersion fibres and
Accepted 3 August 2016
minibars) in enhancing the mechanical behaviour of concrete. Concrete specimens were cast with three
Available online 10 August 2016
different quantities of each fibre, then evaluated based on flexural and drop-weight impact testing.
Interfacial properties were also investigated by scanning electron microscopy. The results indicated both
Keywords:
types of fibre increased pre-cracking strength, but only minibars enhanced the post-cracking behaviour,
Basalt fibre
Minibar
likely due to protection from the polymer.
Basalt fibre reinforced concrete Ó 2016 Elsevier Ltd. All rights reserved.
Post-cracking behaviour
Impact strength

1. Introduction Glass fibre reinforced concrete (GFRC) has been used extensively
to produce thin, light-weight architectural elements, most notably
Plain concrete (PC) is a brittle material with low tensile exterior facade panels. The drawback of GFRC is its susceptibility to
strength. Consequently, PC is susceptible to cracking under tensile degradation in the alkaline environment of concrete, and conse-
stress. When mixed into concrete, randomly distributed fibres are quently, has been largely limited to architectural applications.
able to bridge these cracks and arrest their development. By this It should be noted FRC made with a variety of natural and synthetic
mechanism, it has been well established that the addition of fibres fibres, including carbon, aramid, polypropylene, and wood
can enhance the mechanical behaviour of PC. Although a variety of fibres has been shown to exhibit similar enhancements to the
fibre reinforcing materials exist, fibre reinforced concrete (FRC) mechanical behaviour of concrete [1], though these fibres are not
used for structural applications is most often made with steel currently used as commonly as steel and glass fibres in practical
fibres. The most beneficial properties of steel fibre reinforced applications.
concrete (SFRC) are improved flexural toughness, flexural fatigue Basalt fibre has recently gained popularity as a potential com-
endurance, and impact resistance [1]. As a result, steel fibres are petitor in concrete reinforcing applications due to its excellent
able to totally or partially replace traditional steel rebar in select mechanical properties and an environmentally friendly manufac-
applications, such as industrial floors and pavements. However, turing process [2]. The fibres typically have a tensile strength
SFRC poses several issues, such as: increased dead-load, reduced slightly higher than E-glass fibres, and many times greater than
workability, fibre balling at high dosages, and susceptibility to steel fibres. In addition to plain chopped basalt fibres (BF), a new
corrosion. For these reasons, glass fibre is a popular alternative. basalt concrete reinforcement product called minibars (MB) has
recently been developed. The minibars are essentially a scaled
⇑ Corresponding author. down version of basalt fibre reinforced polymer rebar.
E-mail address: [email protected] (S. Das).

http://dx.doi.org/10.1016/j.conbuildmat.2016.08.009
0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886 879

The research into basalt fibre reinforced concrete (BFRC) has The purpose of the experimental work presented in this paper is
largely been focused on fundamental mechanical properties: com- to compare the pre- and post-cracking mechanical behaviour of
pressive, split-tensile, and flexural strength. In the case of BF, the concrete reinforced with plain chopped basalt fibres (BF), basalt
research shows general agreement with the addition of fibres minibars (MB), and commonly used hooked end steel fibres (SF).
being beneficial up to approximately 0.3–0.5% by volume and Comparative performance is evaluated by flexural and drop-
detrimental thereafter [3–5]. However, optimum fibre dosages weight impact testing. Additionally, interfacial properties are
vary significantly in different types of concrete, such as geopoly- investigated by scanning electron microscopy. It should be noted
mer [6] and high-strength concretes [7]. By comparison, MB have that two types of plain chopped BF are available: filament disper-
been shown to be beneficial at dosages up to 4% by volume [8]. sion and bundle dispersion. Bundle dispersion fibres are manufac-
The influence of BF and MB on compressive strength is typically tured with a sizing that holds bundles of basalt filaments together
not significant [3,5–10], although it has been shown to increase during mixing, whereas filament dispersion fibres will disperse
by as much as 31% with filament dispersion fibres [4]. Literature into individual filaments. In this study, bundle dispersion fibres
suggests the primary benefit of BF and MB in concrete under com- were selected since filament fibres are typically used for crack
pression is the shift from a brittle failure mode to a more ductile control.
one [5,7,8,10].
It has been shown that both BF and MB can significantly 2. Experimental procedure
increase the tensile strength of concrete [3–9]. However, it is diffi-
cult to assess the magnitude of the increase in tensile strength 2.1. Materials
because of discrepancies in values derived from direct tension,
split-tensile, and flexural tests. An increase of 43% in direct tensile All concrete was made with type 10 general use Portland
strength was found using BF with added zirconia, in comparison to cement conforming to the Canadian standard CSA A3001 [16], reg-
a 14% increase without zirconia [9]. Zirconia is added to E-glass ular drinking water, and well-graded aggregates purchased locally.
fibre to produce alkaline resistant glass fibre. This may suggest that Superplasticizer was used in higher dosage FRC mixes.
the BF is susceptible to a similar mechanism of degradation as glass Two different lengths of chopped BF were used: 36 mm and
fibre in concrete. Moreover, Jiang et al. [5] found the beneficial 50 mm. The BF bundles are flat, approximately 0.6 mm wide and
effects of BF diminished significantly after 90 days. made of 16 lm diameter filaments. The MB used in this study
Research related to characterizing the post-cracking perfor- are an epoxy based resin reinforced with 17 lm diameter basalt fil-
mance of BFRC has been limited. This is a problem because in many aments. The composite is 43 mm in length and approximately
practical applications, first-crack strength is not increased. Rather, 0.65 mm in diameter. By comparison with the BF, the MB are more
the most significant enhancement from the addition of fibres is the rigid. The SF used in this study are 38 mm in length, 0.9 mm in
post-cracking response [1]. Both BF and MB have been shown to diameter and have hooked ends. The fibres used in this study are
enhance the flexural toughness of concrete [5,6,8,10]. However, it shown in Fig. 1.
is difficult to assess the relative merit of each product since results
are based on different test methods. It was found using the ACI 2.2. Concrete mix design
Committee 544 recommended drop-weight test for impact resis-
tance [11] that BF can significantly enhance performance after Concrete specimens used in this study were cast with a 0.5
cracking [10]. However, the conclusion is based on data from four water-cement ratio and proportions of 1:1.4:2.8 by mass of
or six specimens per concrete mix. This test method is notorious cement, fine aggregate, and coarse aggregate. The cement was gen-
for large variations, requiring approximately 40 specimens per eral use Portland limestone (type GUL), the coarse aggregate had a
mix to keep the percent error of measured mean values below maximum size of 19 mm, and the fine aggregate had a fineness
10% [12,13]. Li and Xu [14] found BF can significantly increase modulus of 2.7. Three different dosages were used for each type
the energy absorption capacity of geopolymer concrete under of basalt fibre reinforcement, ranging from a low dosage to the
impact loading by using a split-Hopkinson pressure bar system. maximum mixable dosage. Despite the use of superplasticizer, it
However, the performance of BFRC under impact in general is was found that dosages beyond 12 kg/m3 and 40 kg/m3 for BF
not well characterized. Since impact test results obtained by differ- and MB, respectively, led to fibre balling and difficulty achieving
ent test methods are generally not comparable [15], the results proper consolidation. A summary of the mix types used in this
from a simple test method may provide a more practical reference work is shown in Table 1. Mix designation is labelled according
for which future comparisons can be made. This is particularly use- to fibre type, fibre length, and dosage. For example, mix designa-
ful for BFRC because it is a relatively new composite and further tion BF-36-8 indicates chopped basalt bundle dispersion fibres of
development can be expected to enhance its material properties 36 mm length were used at a dosage of 8 kg per 1 m3 of plain
for concrete reinforcing applications. concrete (8 kg/m3).

(a) BF (b) MB (c) SF

Fig. 1. Fibres used in experimental work.
880 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886

Table 1
Test Matrix.

Specimen Fibre Type Length Dosage Plasticizer

(mm) 3 (mL)
kg/m Volume (%)
PC No fiber 0 0 0
BF 36-4 Bundle dispersion 36 mm 4 0.15 0
BF 36-8 8 0.31 30
BF 36-12 12 0.46 60
BF 50-4 50 mm 4 0.15 0
BF 50-8 8 0.31 30
BF 50-12 12 0.46 100
MB 43-6 Minibar 43 mm 6.2 0.31 0
MB 43-20 20 1 0
MB 43-40 40 2 0
SF 38-40 Steel 38 mm 40 0.51 0

2.3. Test methods

Compressive strength was evaluated in accordance with ASTM

C39 [17], using five 100 mm diameter by 200 mm tall cylinders
per mix designation.
Flexural testing was completed following the guidelines of
ASTM C1609 [18]. Concrete prisms 610 mm in length and
152 mm by 152 mm in cross-section were subjected to third-
point loading using a compression testing machine with a
2200 kN capacity. Mid-span deflection was measured using a
25 mm linear displacement transducer (LDT). Mean values
Fig. 3. Impact test fixture.
reported for each mix designation are based on three specimens
tested after 28 days of curing. The test setup is shown in Fig. 2.
Impact resistance was evaluated using a modified version of the
ACI Committee 544 recommended drop-weight impact test [11], as
recommended by Badr and Ashour [12]. Concrete was cast in stan-
dard 152 mm diameter by 305 mm cylindrical moulds with
25.4 mm triangular pieces of wood attached to each side to form
notches. Test specimens 51 mm in thickness were cut from the
notched cylinders using a diamond blade saw. The number of
blows from a 4.54 kg compaction hammer with a 457 mm
(18 in.) drop required to cause a visible surface crack and subse-
quent failure were recorded for each specimen. Failure was defined
by either complete separation of the specimen, separation such
that the specimen is touching both sides of the fixture, or the
impact piston was fully embedded in the concrete. Moreover, only
specimens that cracked through a line between the notches were
included in the data. Mean values reported for each mix designa-
tion are based on 24 specimens tested after 28 days of curing.
Fig. 4. Close-up of impact piston and typical first crack.

The number of specimens tested was based on the statistical

analysis of other researchers using this method [12,19]. The test
setup is shown in Figs. 3 and 4.

3. Results and discussion

3.1. Compressive strength

The compressive strength for each mix designation is shown in

Table 2.
Compressive strength testing was undertaken as a simple
method of evaluating the quality of the concrete. Hypothesis
testing based on a two-tailed t-test indicates it is unlikely that
either BF-36 or BF-50 are useful for enhancing compressive
strength. In fact, the results highlight the importance of proper
specimen preparation to ensure addition of fibres does not result
Fig. 2. Flexural test setup. in too drastic of a decrease in compressive strength, such as that
 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886 881

40
Table 2
Compressive Strength. 35
Specimen No. Compressive COV (%) D COV (%) PC
Samples Strength (MPa)
30
BF-50-12
PC 5 37.90 3.71 –

Load (kN)
25
BF 36-4 5 34.21* 3.60 0.11
BF 36-8 5 39.50 3.19 0.53 20
BF 36-12 5 37.83 5.46 1.75
BF 50-4 5 38.41 4.34 0.63 15
BF 50-8 5 38.96 5.96 2.25
BF 50-12 5 38.68 5.79 2.08 10
MB 43-6 4 23.09* 6.99 3.28
MB 43-20 4 20.90* 7.56 3.85 5
MB 43-40 4 24.84* 6.64 2.93
SF 38-40 5 35.90* 3.20 0.51 0
0 0.2 0.4 0.6 0.8 1
Note:
*
Deflection (mm)
Indicates difference is significant at 95% confidence interval.
Fig. 6. Typical flexural test results for BF-50 specimens.

observed in the case of BF-36-4 and all MB specimens. Using PC as

a benchmark, it can be found from Table 2 that the coefficient of
variation (COV) increases when the BF length increases from 80
36 mm to 50 mm, and is generally higher in MB mixes. This could PC
indicate proper consolidation is more difficult to achieve with 70
MB-43-6
longer fibres and minibars.
Although the addition of MB resulted in a drastic decrease in 60 MB-43-40
compressive strength, they also noticeably enhanced the tough- Load (kN)
50
ness of the concrete under compression. With the exception of
SF, all other specimens could be broken easily apart by hand after 40
removal from the testing machine. Following cracking at peak
compressive load, the MB specimens could be subject to signifi- 30
cantly more compressive load from the testing machine before
20
fracturing. However, this aspect of behaviour was not formally
measured and will not be further discussed. In a more general 10
sense, it is evident the addition of basalt fibres would not be an
effective method of increasing compressive strength. Rather, it is 0
0 2 4 6 8 10
more important to ensure their addition does not hinder compres- Deflection (mm)
sive strength, and that adequate consolidation can be achieved in
the field. Fig. 7. Typical flexural test results for MB-43 specimens.

3.2. Flexural testing and ductility. Three distinct types of failure were observed in MB
specimens depending on fibre dosage. At a low dosage (MB-43-6),
It can be found from the load-deflection plot in Figs. 5 and 6 that the load capacity dropped after the concrete cracked, then increased
BF specimens did not enhance post-cracking behaviour. BF was not again but remained below the first-peak load. At an intermediate
observed bridging the cracks during testing and the specimens dosage (MB-43-20), the load capacity also dropped when the
failed in the same brittle manner as PC. Conversely, Fig. 7 shows concrete cracked; however, it quickly regained and increased beyond
that MB specimens provided substantial post-cracking strength the first-peak load. At a high dosage (MB-43-40), it was unclear when
the concrete first cracked since sudden drop in the load was not

40

35
Table 3
PC Flexural testing results.
30
Mix f1 d1 fp dp fL/600 fL/150 RL/150
25 BF-36-12
Designation (MPa) (mm) (MPa) (mm) (MPa) (MPa) (%)
Load (kN)

20 PC 4.05 0.32 0.15

BF-36-4 4.39 0.37 1.19
15 BF-36-8 4.70* 0.23 0.12
BF-36-12 4.93* 0.33 0.42
10 BF-50-4 4.40* 0.28
BF-50-8 4.89* 0.26
BF-50-12 5.11* 0.27 0.31
5
MB-43-6 4.01 0.17 2.12 2.11 54.38
MB-43-20 3.42 0.17 6.14* 1.30 4.98 5.69 92.77
0
MB-43-40 2.63 0.33 9.22* 2.00* 7.54 8.66 93.74
0 0.2 0.4 0.6 0.8 1
SF-38-40 5.28* 0.60 3.33 1.65 32.79
Deflection (mm)
Note:
*
Fig. 5. Typical flexural test results for BF-36 specimens. Indicates difference is significant at 95% confidence interval.
882 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886

(a) BF-36-12 (b) MB-43-40

Fig. 8. Cracked cross-section of failed flexural test specimens.

observed at any point and the load-deflection plot followed a smooth Table 4
softening behaviour after reaching peak-load. Impact test results.
All the results are summarized in Table 3. In this table, f1 is the Mix 152 mm Drop 457 mm Drop % Change
first-peak stress value, d1 is the deflection at first-peak stress, fp is Designation
N1 N2 N1 N2 N1 N2
the maximum stress, and dp is the deflection at maximum stress.
The fL/600 is the residual strength at the deflection of L/600 and PC 5.0 3.2 1.0
BF-36-4 3.9 2.4 23.1 24.7*
fL/150 is the residual strength at the deflection of L/150, where L BF-36-8 4.0 3.0 19.8 5.2
is the span length of the beam. The RL/150 is the flexural strength BF-36-12 4.3 2.4 14.9 24.7*
ratio calculated as per ASTM 1609 [18]. BF-50-4 4.9 3.8 2.5 16.9
The poor post-crack performance of BF (Figs. 5 and 6) was not BF-50-8 5.7 4.2 12.4 29.9*
BF-50-12 4.4 3.3 13.2 2.6
surprising given the lack of visible fibres in the cracked cross-
MB-43-6 5.6 13.3 11.6 1233.3*
section, as shown in Fig. 8. Even at the highest dosages, BF was MB-43-20 8.0 19.4 59.5* 1837.5*
not visible by eye, and thus, some level of degradation in the MB-43-40 9.3 30.6 84.3* 2962.5*
chopped fibres was suspected. Regardless, the BF increased the SF-38-40 6.7 19.7 32.2* 1870.8*
first-peak stress, and therefore, provided some benefit. In most Note:
cases, the deflection at the first-peak stress was found to be lower *
Indicates difference is significant at 95% confidence interval.
in BF specimens in comparison with PC (Figs. 5 and 6 and Table 3).
This would suggest the influence of BF is an increase in strength
and modulus of elasticity as fibre dosage increases. Moreover, this
increase is greater with fibres 50 mm in length than with fibres
Statistical analysis was completed with the Mann-Whitney
36 mm in length. The MB specimens were also effective in increas-
U-test, since it was unclear if the data followed a normal distribu-
ing first-peak load (Fig. 7). In this regard, MB-43-20 specimens pro-
tion. The analysis shows BF does not have a statistically significant
vided a similar increase as the SF-38-40 specimens. Furthermore,
effect on first-crack strength (N1), and only in a few cases was
MB specimens behaved in a ductile manner after cracking. As a
found to significantly affect post-cracking performance (N2). At
result, the MB specimens were able to carry between 50% and
best, BF-50-8 was found to increase N2 by approximately 30%.
90% of peak-load at a deflection of 3 mm (L/150) and still had
On the other hand, BF-36-4 and BF-36-12 were found to decrease
residual load capacity at a large deflection of 10 mm. This is
N2 by approximately 25%. It is believed these differences are the
because the MB composites failed primarily by gradual fibre pull-
result of an insufficient sample size, since unlike the MB mixes,
out, evidenced by the lack of ruptured fibres observed in the
there was no obvious trend in the data and the BF was not visible
cracked cross-section (Fig. 8). The post-cracking performance of
bridging the crack. All of the data is presented graphically in Figs. 9
MB-43-6 specimens was comparable to that of SF-38-40 (Table 3).
and 10. In these graphs, the ultimate resistance (sum of N1 and N2)
is depicted. The error bars shown represent one standard deviation
3.3. Impact testing on either side of the measured mean. Based on the size and overlap
of the error bars, it can be seen visually from Fig. 9 that the
It was found after several preliminary tests that all 24 PC spec- differences in the mean values between PC and BF specimens are
imens cracked after a single blow and failed after one additional likely not significant. On the other hand, the differences in mean
blow, if not already failed, when using the full drop height of the values in the MB specimens are clearly significant (Fig. 10).
hammer (457 mm). A similar performance was observed with five The cracked cross-sections shown in Fig. 11 are similar to those
specimens from all other mix designations (Table 1). Therefore, in in Fig. 8, in which fibres are not visible in BF specimens, while
the subsequent impact tests, the drop-height was reduced to fibres in MB specimens are clearly effective preventing separation
152 mm (6 in.). However, with the reduced drop-height of by bridging the crack.
152 mm, the MB-43 and SF-38 specimens required above 100 An obvious trend in the data is present, where increasing
blows to fail, which was deemed impractical. Thus, after the first dosages of MB resulted in increases in N2 ranging from approxi-
crack was observed in SF-38 and MB-43 specimens, the hammer mately 1200–3000% (Table 4). MB-43-20 specimens had a similar
was dropped from the full height (457 mm) and compared with performance to SF-38-40 specimens, with an increase in post-
PC subjected to the same impact. Table 4 shows the mean number cracking impact strength (N2) of approximately 1900%. In both
of blows until the first crack (N1) and subsequent number of blows MB and SF specimens, observation of the cracked cross-section
until failure (N2). showed nearly all fibres failed by pull-out. Ruptured fibres were
 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886 883

variation of the ultimate resistance (sum of N1 and N2) ranged

from 19.2 to 40.3%, and was typically 25–30%. These values are
similar to those reported by other researchers using the same test
method [12,19]. This high variation in behaviour could deter prac-
tical application of the fibre, since it is probable that a practicing
engineer would prefer to have a higher level of certainty in
expected behaviour. Additionally, the results are confined to the
loading conditions of the test method (4.54 kg dropped from
152 mm to 457 mm). The effect of minor changes to the drop
height, as discussed at the start of the chapter, suggest mechanical
behaviour would likely change drastically when subjected to
different loading rates and magnitudes, such as explosive loads,
vehicle collisions, or airplane landings. It is the opinion of the
authors that these are the two biggest limitations of the results,
and consequently, project specific testing would be mandatory
Fig. 9. Impact test results of BF specimens. for successful application of the fibres. However, the results
provide a useful benchmark for on-going testing as the material
properties of the fibres are uncovered for optimization for concrete
reinforcing applications.

3.4. Interfacial properties

The scanning electron microscope (SEM) image in Fig. 12(a)

shows the penetration of cement hydration products in between
individual filaments of a BF bundle. This could explain the brittle
nature of the BF composites observed in this work, whereby the
growth of hydration products between the filaments increases
the fibre-matrix bond strength beyond the tensile strength of the
fibres, resulting in failure governed by fibre rupture. This is further
evidenced by the lack of visible fibres in the failed cross-sections,
as shown in Figs. 8(a) and 11(a). Moreover, deflection at peak-
load is generally lower in BF-50 specimens than BF-36 specimens
Fig. 10. Impact test results of MB and SF specimens. (Table 3). Thus, failure may be a combination of pull-out and rup-
ture. In both cases, failure would be due to fibre rupture; however
the 50 mm fibres probably slipped a little less than the 36 mm
more prevalent in SF specimens, which is likely due to the increase fibres due to increased bond strength from a greater contact area
in bond strength from the hooked ends of the SF. Moreover, MB with the matrix. The same phenomenon has been reported in liter-
and SF were also effective in increasing the first-crack strength ature related to GFRC [20–22], though it is not agreed upon if the
of the concrete (N1). The increase in first-crack strength of composites lose toughness primarily due to a physical mechanism,
MB-43-20 specimens was similar to that of SF-38-40 specimens. as suggested in this case, or by a chemical mechanism. In the case
Although it should be noted that using the full drop-height of MB specimens, the polymer appeared to be effective in prevent-
(457 mm) during preliminary testing, the MB and SF specimens ing the penetration of hydration products. Some cracks were
cracked after no more than two blows. This would suggest basalt observed in the polymer, as shown outlined in a white broken line
fibre reinforcement does not have a meaningful effect on first- in Fig. 12(b). The cracks are believed to be the result of mechanical
crack strength under impact loading; it is only useful in enhancing damage during mixing, though may also be due to shrinkage stress,
the post-cracking behaviour. manufacturing defects, or transportation. Pull-out failure observed
It is evident from the standard deviation bars in Figs. 9 and 10 during testing would indicate it is unlikely the cracks have a
that there was large variation in the data. The coefficient of significant influence on 28 day performance of impact and flexural

(a) BF-50-12 (b) MB-43-20

Fig. 11. Cracked cross-section of failed impact test specimens.
884 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886

Crack

(a) Bundle dispersion fibre (b) Minibar

Fig. 12. SEM images of fibres in concrete after 7 days.

testing. If the cracks were detrimental it would be expected that

more of the fibres would fail by rupturing instead of pulling out.
However, this may become a durability issue over a longer dura-
tion, or in harsher mixing conditions.
Ongoing work at the University of Windsor has shown BF bun-
dles will abrade more severely during mixing in the presence of
higher quantities of coarse aggregate. Fig. 13 shows the typical
appearance of the cracked cross-section of a BF-50-12 impact spec-
imen. Not only could no fibres be found oriented in a manner that
would suggest they were effectively bridging the crack, but the
bundles had also clearly been separated into individual filaments.
It is believed this is the result of a combination of abrasion during
mixing and the ongoing growth of hydration products between the
filaments. The work of Bentur [20] found that in brittle composites,
spaces between filaments were at least partially filled with hydra-
tion products, while in the case of ductile composites, these spaces
were largely empty. Therefore, future research should address how
to mitigate this issue since ductility is a very desirable trait of FRC.
Jiang et al. [5] observed that the increases in compressive and
flexural strength of BFRC diminished over time and attributed it
to fibres de-bonding from the matrix. The hypothesis was based
Fig. 13. Cracked cross-section of BF-50-12 impact specimen after 9 months. on the development of spaces between the fibre and matrix and
a decrease in the density of cement on the fibre surface between

(a) Aer 7 days (b) Aer 9 months

Fig. 14. Change in cement density on fibre surface over time.
 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886 885

(a) Plain fibre (b) Aer 9 months

Fig. 15. Change in fibre surface after immersion in concrete.

7 and 28 days. In some instances spaces were observed between 1. Fibre dosages beyond 12 kg/m3 and 40 kg/m3 of BF and MB,
the fibres and matrix, but it was believed to be the result of respectively, led to mixing problems due to fibre balling and
mechanical disruption due to testing or specimen preparation. resulted in difficulty handling, placing, and consolidating fresh
Changes in cement density on the fibre surface after 7 and 28 days concrete.
were not obvious. However, distinct differences were observed 2. The addition of BF increased the first-crack strength of concrete
after 9 months. The differences are characterized well by Figs. 14 subjected to flexural loading, but was not significantly influen-
and 15. It can be found in Fig. 14 the amount of cement on the fibre tial when subjected to impact loading. In the case of flexural
surface appears to decrease. Additionally, as shown in Fig. 15, the loading, the first-crack strength increases with increasing fibre
roughened fibre surface after 9 months indicates a chemical dosage. The strength improvement was greater using longer,
reaction may have taken place. 50 mm BF than with 36 mm BF. A dosage of 12 kg/m3 of
Scheffler et al. [23] have shown BF corrodes in a cement solu- 50 mm BF resulted in a first-crack strength that was compara-
tion, characterized by the development of small holes on the fibre ble to a dosage of 40 kg/m3 of SF.
surface after 7 days. Previous research has also shown basalt fibres 3. The addition of MB increased the first-crack strength of con-
will lose tensile strength over time when immersed in a solution of crete subjected to both flexural and impact loading. In both
calcium hydroxide intended to replicate a hydrating cement cases, the first-crack strength increased with increasing fibre
medium [24,25]. Therefore, the relative poor performance of BF dosage. However, at higher fibre dosages it was difficult to
observed in this work is believed to be analogous to the well- assess when the concrete cracked since the composite behaved
established aging process of GFRC. The aging process of GFRC has in a ductile manner. A dosage of 20 kg/m3 of MB resulted in a
two primary mechanisms: firstly, a physical mechanism character- comparable increase in first-crack strength to that of SF with
ized by the growth of hydration products between the filaments, a dosage of 40 kg/m3.
and secondly, a chemical attack due to the high alkalinity of the 4. The addition of BF at any dosage did not have a meaningful
cement matrix [26]. This is unsurprising given the similar manu- effect on the post-cracking behaviour of concrete. On the other
facturing process and chemical composition of basalt and glass hand, MB had a significant benefit, which was further enhanced
fibres. This research shows the use of a polymer is effective in over- with increasing fibre dosages. Fibre dosages of 6 kg/m3 and
coming these problems, and thus, MB are a promising alternative 20 kg/m3 of MB resulted in a comparable post-cracking perfor-
to steel fibres for concrete reinforcement. Although some research mance to SF at a dosage of 40 kg/m3 under flexural and impact
has been done to quantify the long-term durability of BFRP rebar loading, respectively.
[27], similar work should be undertaken for MB due to the substan- 5. The poor post-cracking response of BF specimens was attribu-
tial increase in surface area and the potential for damage during ted to failure by fibre rupture, in comparison to the MB
mixing. Finally, it should be noted that a vast amount of research specimens which failed primarily by fibre pull-out. The ductile
exists on mitigating the aforementioned problems with GFRC by post-cracking behaviour in MB specimens was the result of
the addition of pozzolanic fillers to ordinary Portland cement con- failure by gradual pull-out.
crete mixes, or with the use of alternative cements [26]. This would 6. Cement hydration products were observed in between the indi-
likely explain why Dias and Thaumaturgo [6] found BF performed vidual filaments of the bundles of BF. Moreover, physical
better in a geopolymer concrete than in Portland cement concrete. changes to the BF surface were observed after nine months
Future research into BFRC can likely be expedited by taking and believed to be indicative of degradation to mechanical
advantage of the parallels drawn with GFRC and the enormous properties. Hence, the brittle behaviour of BF composites can
amount of previous work already published in that field. be attributed to fibre rupture due to a combination of increased
fibre-matrix bond strength and decreased fibre tensile strength.

4. Conclusions
Acknowledgements
The conclusions presented in this paper are based on the results
obtained from this research. Hence, the following results presented The authors appreciate the financial assistance received from
in this paper may be limited to the specimens used in this study: OCE, NSERC, Connect Canada, and MEDA Limited. The authors also
886 J. Branston et al. / Construction and Building Materials 124 (2016) 878–886

appreciate the technical assistance and donation of materials from [13] M. Nataraja, N. Dhang, A. Gupta, Statistical variations in impact resistance of
steel fiber-reinforced concrete subjected to drop weight test, Cem. Concr. Res.
MEDA Limited.
29 (7) (1999) 989–995.
[14] W. Li, J. Xu, Mechanical properties of basalt fiber reinforced geopolymeric
concrete under impact loading, Mater. Sci. Eng. A 505 (1) (2009) 178–186.
References [15] S. Mindess, Thirty years of fibre reinforced concrete research at the University
of British Columbia, in: Proceedings of First International Conference on
[1] ACI Committee 544, 544.1R-96: Report on Fiber Reinforced Concrete Sustainable Construction Materials and Technologies, 2007, pp. 259–268.
(Reapproved 2009). Technical Documents, 1996. [16] CSA, Cementitious Materials Used in Concrete, 2013. Mississauga, Ontario.
[2] A. Ross, Basalt Fiber: Alternative to Glass? wwwcompositesworld.com/articles/ [17] ASTM, Standard Test Method for Compressive Strength of Cylindrical Concrete
basalt-fibers-alternative-to-glass2006 (accessed 17.09.2015). Specimens. ASTM C39/C39M-05, 2006. West Conshohocken Pennsylvania.
[3] T.M. Borhan, Thermal and mechanical properties of basalt fibre reinforced [18] ASTM, Standard Test Method for Flexural Performance of Fiber Reinforced
concrete, in: Proceedings of World Academy of Science, Engineering and Concrete (using Beam with Third-point Loading). ASTM C1609/C1609M-05,
Technology, World Academy of Science, Engineering and Technology (WASET), 2006. West Conshohocken Pennsylvania.
2013, pp. 712–715. [19] J.J. Myers, M. Tinsley, Impact resistance of blast mitigation material using
[4] P. Iyer, S.Y. Kenno, S. Das, Mechanical properties of fiber-reinforced concrete modified ACI drop-weight impact test, ACI Mater. J. 110 (3) (2013) 339–348.
made with basalt filament fibers, J. Mater. Civ. Eng. (2015). 04015015. [20] A. Bentur, M. Ben-Bassat, D. Schneider, Durability of glass-fiber-reinforced
[5] C. Jiang, K. Fan, F. Wu, D. Chen, Experimental study on the mechanical cements with different alkali-resistant glass fibers, J. Am. Ceram. Soc. 68 (4)
properties and microstructure of chopped basalt fibre reinforced concrete, (1985) 203–208.
Mater. Des. 58 (2014) 187–193. [21] A. Peled, J. Jones, S. Shah, Effect of matrix modification on durability of glass
[6] D.P. Dias, C. Thaumaturgo, Fracture toughness of geopolymeric concretes fiber reinforced cement composites, Mater. Struct. 38 (2) (2005) 163–171.
reinforced with basalt fibers, Cement Concr. Compos. 27 (1) (2005) 49–54. [22] P. Purnell, N.R. Short, C.L. Page, A.J. Majumdar, Microstructural observations in
[7] T. Ayub, N. Shafiq, M.F. Nuruddin, Mechanical properties of high-performance new matrix glass fibre reinforced cement, Cem. Concr. Res. 30 (11) (2000)
concrete reinforced with basalt fibers, Proc. Eng. 77 (2014) 131–139. 1747–1753.
[8] S. Adhikari, Mechanical and structural characterization of mini-bar reinforced [23] C. Scheffler, T. Förster, E. Mäder, G. Heinrich, S. Hempel, V. Mechtcherine, Aging
concrete beams (Ph.D. dissertation), The University of Akron, 2013. of alkali-resistant glass and basalt fibers in alkaline solutions: evaluation of the
[9] Y.V. Lipatov, S. Gutnikov, M. Manylov, E. Zhukovskaya, B. Lazoryak, High failure stress by Weibull distribution function, J. Non-Cryst. Solids 355 (52)
alkali-resistant basalt fiber for reinforcing concrete, Mater. Des. 73 (2015) (2009) 2588–2595.
60–66. [24] J.J. Lee, J. Song, H. Kim, Chemical stability of basalt fiber in alkaline solution,
[10] V. Ramakrishnan, N.S. Tolmare, V. Brik, Performance evaluation of 3-D basalt Fibers Polym. 15 (11) (2014) 2329–2334.
fiber reinforced concrete & basalt rod reinforced concrete, Innovations [25] F. Rabinovich, V. Zueva, L. Makeeva, Stability of basalt fibers in a medium of
Deserving Exploratory Analysis Programs (IDEA) Program Final Report, hydrating cement, Glass Ceram. 58 (11–12) (2001) 431–434.
pp. 1–79. [26] A. Bentur, S. Mindess, Fibre Reinforced Cementitious Composites, CRC Press,
[11] ACI Committee 544, Measurement of properties of fiber reinforced concrete, 2006.
ACI Mater. J. 85 (6) (1988). [27] A. Serbescu, M. Guadagnini, K. Pilakoutas, Mechanical characterization of
[12] A. Badr, A.F. Ashour, Modified ACI drop-weight impact test for concrete, ACI basalt FRP rebars and long-term strength predictive model, J. Compos. Constr.
Mater. J. 102 (4) (2005) 249–255. 19 (2) (2014) 04014037.