Composite Structures 183 (2018) 389–396

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Experimental study of concrete beams prestressed with basalt fiber

reinforced polymers under cyclic load
Edgaras Atutis ⇑, Juozas Valivonis, Mantas Atutis
Vilnius Gediminas Technical University (VGTU), Sauletekio Ave. 11, Vilnius 10223, Lithuania

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

Article history: This paper summarizes a study undertaken to analyse the fatigue behaviour and variation on modulus of
Received 30 December 2016 elasticity of basalt fiber reinforced polymer (hereinafter-BFRP) under experimental cyclic loading. The
Revised 22 February 2017 impact of load frequency and amplitude of cyclic load to ultimate strength of BFRP and the number of
Accepted 29 March 2017
the cycles to degradation of modulus of the elasticity were investigated, respectively. Based on experi-
Available online 8 April 2017
mental results a relationship of stress range and number of load cycles of BFRP bars is proposed. Also,
brand new results of non-common fatigue resistance experiments of concrete beams prestressed by
Keywords:
BFRP bars are introduced. Moreover, impact of the number of load cycles to deflection and crack width
Prestressed concrete
Cyclic load
is estimated. The effect of the degree of the prestressing to concrete beams under cyclic load is described
basalt fiber-reinforced polymer (BFRP) and the main figures and relationships appear in the text below.
Deflection Ó 2017 Elsevier Ltd. All rights reserved.
Crack
Fatigue

1. Introduction possible until a fundamental understanding of the fatigue resis-

Composites use dates back to the earliest civilizations, but the
use of engineered composites dramatically increased in the last
century. Codes and standards are rapidly evolving to keep up with
the recent growth of composites in civil infrastructure such as
bridges, parking slabs, industrial facilities, airplane components,
wind turbines, high pressure pipes, natural gas transmission etc.
Advances in the fields of fiber reinforced polymers (hereinafter-
FRPs) have resulted in the development of high strength, light-
weight and cost-effective design life. Also, FRPs are characterized
by excellent corrosion resistance, fatigue resistance and low den-
sity, thus the use of such high performance materials might
increase the flexural strength and stiffness of concrete elements
and improved durability over conventional steel either the external
(strengthening) or internal reinforcement would be employed [1–
3]. Among these, the application of FRP to strengthen concrete
beams perhaps received the most attention from the research com-
munity and many researchers have reported improvements in
strength and stiffness of strengthened beams [4–6]. As mentioned
above, FRP offer a great potential for fatigue resistance and further
research is significantly needed in order to expand application of
composite materials in the industry practice [7]. Nevertheless,
effective application of FRP to structural concrete systems is not
⇑ Corresponding author.
E-mail address: edgaras.atutis@vgtu.lt (E. Atutis).
tance and failure mechanisms under cyclic load is available: fati-
gue behaviour of FRP bars in air and surrounded by concrete [8],
more reliable, innovative and efficient anchorage systems are
introduced for reinforced concrete applications [9–12], fatigue life
of different type of structures undergoing cyclic loading [13–16].
With respect to fatigue of structural element, attention shall be
addressed to degradation of the mechanical characteristics of the
certain material, where the stress depend on load amplitude under
repeat loading. Moreover, those stress are varied due to time and
considering fatigue resistance of the structure a long-term test,
in most cases, are being employed. In order to understand the
mechanism of the behaviour of the structural element under cyclic
loading, separate analysis of certain materials, forming that ele-
ment, is required. Concrete stress-strain relationship depends on
impact of cyclic load as well. The effect of fatigue in the concrete
is caused due to variation of plastic deformation, thus, cracking
occurs and vibro-creep (hereinafter-fatigue creep) appears. In
addition, if concrete stresses are increased by cyclic load, the effect
of fatigue creep is even more significant and notable, and thus,
growth of cracking occurs due to damage of the bond interface
between concrete and polymer matrix. It was observed that fatigue
resistance is not effected by cyclic load frequency for concrete
stresses less than 75 percent of the ultimate concrete compressive
strength [17]. This paper first reviews behaviour of FRP bars under
cyclic load and concrete fatigue resistance is not to be the topic of
this research.

http://dx.doi.org/10.1016/j.compstruct.2017.03.106
0263-8223/Ó 2017 Elsevier Ltd. All rights reserved.
390 E. Atutis et al. / Composite Structures 183 (2018) 389–396

Failure due to fatigue of homogenous materials, such as steel,

occurs depending on first major crack and crack growth by increas-
ing the number of the load cycles while failure mechanism of the
composite materials, such as FRP, is different than conventional
steel and mostly consists of the micro-cracking of FRP matrix,
debonding of fibers and matrix, and finally, rupture of the fibers.
Such ‘‘several stage” process appear due to accumulated damages
instead of opening of the first crack and crack growth. In literature,
range of applied stress or strain is most significant and determine
failure of composite materials caused by the fatigue [7,8,14]. Fur-
ther, fatigue resistance of FRP materials shall be described by three
levels of damage mechanisms (see Fig. 1). At the first level (maxi-
mum stress or strain), failure occur due to the rupture of the fibers
and interfacial debonding where greatest deformation exist.
Fig. 2. Degradation of short-term strength of various types of FRPs [19].
Increasing the number of fibers broken, stress and strain pro-
gressively increases in fibers, which are still not damaged and time
period in between decreases in comparison with already ruptured
compression and shear. The tension-tension test method was cho-
fibers until material observe the failure. At lower stress or strain,
sen for this set of the experimental work. The main test control
micro-cracking of matrix and shear of layers (interfacial shear)
parameter was the load. BFRP specimen is subjected to repetitive
are dominant factors. FRP is able to maintain large number of load
constant load range. This test method was intended to develop
cycles at minimum level of stress or strain. It is considered that
the tensile fatigue characteristics of BFRP (fatigue life data). Such
long-term cracks shall not open at this stage and FRP can resist
data was used to create S-N curves for the specimen tested under
stress caused by cyclic load. Moreover, mentioned level of stress
particular testing conditions where the main variable is the maxi-
or strain does not influence failure of the material [16,18].
mum value of the repeated load [20,22]. In order to determine an
During fatigue process short-term strength (hereinafter-fst) or
S-N curve, the maximum and minimum loads shall be set by one
modulus of elasticity of material is decreased. Therefore, failure
of the following methods: fix the average load and vary the load
of material might occur at much lower stress compared to short-
amplitude; fix the minimum load and vary the maximum load;
term strength of FRP. However, composite materials are considered
fix the load ratio and vary the maximum and minimum load
to have great fatigue resistance (hereinafter-ffat). In accordance
according to the fixed ratio, accordingly. In this paper, S-N curve
with literature, regression of fst of FRP, in particular to aramid
for BFRP bars was determined by second method mentioned above.
(hereinafter-AFRP), glass (hereinafter-GFRP) and carbon
The main argument for the selection of the method was based on
(hereinafter-CFRP) fiber reinforced polymers under increased
possibility to evaluate the impact of load frequency and amplitude
number of load cycles (hereinafter-N) appear in Fig. 2 [19]. Yet,
to final number of cycles. Also, the aim was to determine degrada-
not many information about the research of short-term fatigue
tion of modulus of elasticity due to multiple load effect, if any.
strength variation of basalt FRP bars can be found in scientific lit-
BFRP reinforcement, commercially known as RockBar Composite
erature. This led to arrange experimental study in order to predict
bars were employed (see Fig. 3) for the test. Major mechanical
the behaviour of such brand new composites. After separate BFRP
characteristics of BFRP bars were developed from static tension-
fatigue tests (material) results an experimental program and real-
tension tests conducted under the procedure of such test methods
istic flexural behaviour under cyclic load of BFRP prestressed con-
[20,22]. Obtained results are introduced in Table 1 and were used
crete beams were undertaken.
in order to prepare dynamic testing program.
In general, 6 fatigue bar specimens with certain length of
2. Fatigue analysis of BFRP bars 1000 mm and 12.45 mm diameter were prepared. Principal
scheme of fatigue test is shown in Fig. 4. In order to proceed with
In general, at high-cycle fatigue situations, materials perfor- experiment the ends of all specimens were anchored in steel tube
mance is commonly characterized by cyclic stress range
(hereinafter-S) and N curve, also known as Wöhler curve. This is a
graph of the magnitude of S against the logarithmic scale of cycles
to failure. Various types of fatigue testing are possible. Most com-
mon tests are available, in particular to tension-tension; tension-
compression; compression-compression; combinations of tension;

Fig. 3. View of BFRP specimens (Courtesy of the authors).

Table 1
Main characteristics of BFRP at static tension test.

Load type ffu (MPa) efu Ef (MPa)

Static load 1089.9 0.0246 45.0
Fig. 1. Fatigue life for unidirectional composites.
 E. Atutis et al. / Composite Structures 183 (2018) 389–396 391

parameters are being evaluated: rate of deformation of the bar;

time under mean load; and heat generation. These parameters
have different impact to mechanical characteristics of composite
materials [7]. During the test 2 most common failure modes of fati-
gue bar specimens were obtained: splitting failure (see Fig. 5 (b));
and broom-shape failure (see Fig. 5 (c)). The last one occurred due
to rupture of the rest of polymer fibers consisting FRP bar.
Further results of experiment are presented in Fig. 6, where the
maximum applied stress is against the number of cycles are
shown. It appears, that the logarithm of number of the cycles
decreases linearly when the stress range increases. It can be
described by equation bellow:

Sr ¼ A  N B ð1Þ

log Sr ¼ C þ B  log N ð2Þ

where Sr is applied stress range (MPa); N is number of cycles to fail-

ure; and A, B, C are constants depending on the material and test
conditions.

Fig. 4. Scheme of fatigue test.

and epoxy resin inside by keeping 375 mm anchor length. This

length had to guarantee no failure of the specimens at anchoring
zone.
Regarding fatigue test program two groups of specimens were
considered. The first group consist of 3 specimens to be tested at Fig. 5. Fatigue test of BFRP bar specimens: a) prior the test b) splitting failure. c)
a frequency of 4 Hz. The second group consist of 3 fatigue bar spec- broom-shape failure (Courtesy of the authors).
imens to be tested at a frequency of 2 Hz at the critical section until
they failed by fatigue rupture. Tests were performed with mini-
mum constant cycle stress (rmin = const.) of the bar. These stress
observed 55 percent of ultimate tensile strength (hereinafter-ffu)
of BFRP. Meanwhile, maximum stress ranged between 65 and 90
percent of ffu. BFRP bars were tested until they reached 1 million
cycles. In fact, once the specimen reached this range of cycles it
was observed that specimen keeps great resistance of fatigue. Test
program matrix and results are provided in Table 2.
With respect to experiment results it was noted, that multiple
cyclic load which produce cyclic stress from 55 percent to 65 per-
cent of ffu no risk shall be concerned. Moreover, specimens tested
at the minimum and maximum stress ratio (hereinafter-R) equal
to 0.88, were at ‘‘safe zone” and were capable to resist more than
1 million cycles. At less R = 0.75, the fatigue life of BFRP bars
decreased rapidly and the same effect was observed at R = 0.62.
Considering a frequency impact to fatigue life of BFRP 3 major Fig. 6. Sr-N response of BFRP bars.

Table 2
Results of the tests.

Spec. No. Min stress (MPa) Max stress (MPa) Stress range (MPa) Frequency (Hz) Number of cycles
BFRP-D-1-2 606 (55% ffu) 687 (65% ffu) 81 2 >1000,000
BFRP-D-2-2 808 (75% ffu) 202 2 13,647
BFRP-D-3-2 970 (90% ffu) 364 2 18
BFRP-D-4-4 687 (65% ffu) 81 4 >1000,000
BFRP-D-5-4 808 (75% ffu) 202 4 9775
BFRP-D-6-4 970 (90% ffu) 364 4 20
rmin = const. rmax – const. Sr – const. f – const.
392 E. Atutis et al. / Composite Structures 183 (2018) 389–396

Based on experimental results of fatigue BFRP bar specimens a

relationship of stress amplitude (hereinafter-Sr) and N is proposed
as
Sr ¼ 444:48  25:95 ln N ð3Þ

3. Modulus of elasticity of BFRP bars under cyclic loading

In order to determine the impact of cyclic load to modulus of

elasticity experiment was arranged with particular estimation of
deformation of BFRP specimens and variation of deformation due
to increased N. Specimens used to investigate degradation of mod-
ulus of elasticity were identical to those used for fatigue perfor-
mance analysis (see Fig. 4). Deformation was measured at the
critical section (in the middle of the bars). Fatigue bar specimens
were tested at a frequency of 4 Hz at R = 0.88 (see Fig. 7). Moreover,
ratio of measured modulus of elasticity at the certain load cycle
(hereinafter-Ei) and initial modulus of elasticity prior tension-
tension test (hereinafter-E) of BFRP bars, respectively (see Fig. 8).
Two specimens were tested and it was observed, that modulus
of elasticity decreased only by 4 percent after 1 million cycles, thus
it shall be noted, that fatigue BFRP bar specimens perform great
fatigue resistance under cyclic load considering small variation of
the modulus of elasticity.

4. Fatigue analysis of BFRP prestressed concrete beams

This paper, also, summarizes the study undertaken in order to

Fig. 7. View of the test. (Courtesy of the authors).
Fig. 8. Modulus of elasticity variation. curve based on N.
determine flexural behaviour of concrete beams prestressed with
composite bars (hereinafter-PC) under cyclic load. Twelve beams
prestressed by BFRP reinforcement, commercially known as Rock-
Bar Composite were tested. Moreover, all beams were divided into
3 groups by 4 specimens based on the degree of prestressing force.
The beams were 3.2 meter long and rectangular cross section with
150 mm width and 300 mm height. Concrete compressive strength
varied from 40.2 to 46 MPa (see Tables 3 and 4). Geometrical
parameters and test set-up of the specimens appear in Fig. 9. The
three jacking stress levels used in the study 40, 45 and 55 percent
of the ultimate guaranteed strength (hereinafter-ffu) of BFRP. The
level of prestressing and consequently the concrete stress distribu-
tion along the section was varied by using one bar. In order to
avoid the failure in shear beams were reinforced by transversal
8 mm diameter steel reinforcement. Anchorage system based on
steel tube anchors was used for the tests of BFRP bar (Fig. 4). BFRP
bars were prestressed by hydraulic jack and the ends of bars were
anchored at fixed frame supports. After 14 days of concrete hard-
ening bars were released from supports and prestress force

Table 3
Characteristics of beams.

Beam. No. Beam height (mm) Beam width (mm) Span length (mm) ffu (MPa) efu Ef(GPa) fck (MPa) q (%)
Group 1
CSI-1 303 149 2640 1089.9 0.0246 45.0 46.01 0.32
CSI-2 300 148 46.01 0.33
CSI-3 301 150 44.47 0.32
CSI-4 299 149 44.47 0.33
Group 2
CSI-5 300 150 40.21 0.32
CSI-6 303 152 40.21 0.32
CSI-7 301 148 40.21 0.33
CSI-8 303 147 40.21 0.33
Group 3
CSI-9 301 151 44.29 0.32
CSI-10 303 153 44.29 0.31
CSI-11 302 150 44.29 0.32
CSI-12 300 149 44.29 0.33
 E. Atutis et al. / Composite Structures 183 (2018) 389–396 393

Table 4
Load characteristics and fatigue lifes of beams.

Group Beam No. Level of prestress (%) Load level Fmin (kN) Load level Fmax (kN) Load range (kN) Load frequency N
(Hz)
I CSI-1 55 20.0 67.0 47.0 1.0 28,031
CSI-2 55 28,145
CSI-3 40 7345
CSI-4 40 7687
II CSI-5 45 57.0 37.0 82,348
CSI-6 45 86,684
CSI-7 45 47.0 27.0 172,244
CSI-8 45 193,475
III CSI-9 45 53.0 33.0 199,851
CSI-10 45 207,451
CSI-11 45 43.0 23.0 375,263
CSI-12 45 392,159
Fmin = const. f = const.

compressed concrete by transferring compression in the tension
zone. Fatigue tests of prestressed beams were performed starting
with Group 1 specimens 14 days afterdays.
In regard to local laboratory premises (Vilnius Gediminas tech-
nical university) beams were specified by original notations. Letter
‘‘S” refers to the type of member (‘‘Sija” = ‘‘Beam”) and ‘‘CSI” refers
to reinforcing method and type of loading (‘‘sija zitemptoji ban-
doma cikline apkrova” = ‘‘prestressed beam tested by cyclic load”).
Experimental beams were instrumented with strain gauges at
side surface and deflection gauges at the bottom of the beams as
shown in Fig. 9, respectively. Also, linear voltage displacement
transducers (hereinafter-LVDTs) were placed at the BFRP bar loca-
tion of both ends of the beam to monitor any possible slippage dur-
ing the experiment.
Prestressing force of beams CSI-1 and CSI-2 (group 1) was 55
percent of ffu, whereas, beams CSI-3 and CSI-4 (group 1) obtained
15 percent less prestressing force than previous two. All beams
were tested under cyclic load. The minimum and maximum values
of the load were 20 kN and 67 kN, respectively. In regard to maxi-
mum load (67 kN), it approximately corresponded 70 percent of
the ultimate static strength (hereinafter-Fu.static) determined based
on codes and standards [19,21,23]. With respect to beams CSI-5,
CSI-6, CSI-7, CSI-8 (group 2) prestressing force reached 45 percent
of ffu. The minimum and maximum load were 20 kN and 57 kN,
correspondingly. Maximum load was about 60 percent of Fu.static.
Prestressing force of group 3 (CSI-9, CSI-10, CSI-11, CSI-12) speci-
mens obtained 45 percent of ffu. The minimum load was 20 kN
and maximum load reached 55 percent of Fu.static (53kN). All beams
were tested at a frequency of 1 Hz. Example of fatigue beam spec-
imen prior the and during the test is shown in Fig. 10.
The minimum load (20 kN) and maximum load (67 kN) were
constant over the all specimens of group 1. In comparison, beam

Fig. 9. Geometrical parameters, test set-up (four-point loading) scheme and

arrangement of LVDTs. Fig. 10. PC beam a) prior the test b) during the test.
394 E. Atutis et al. / Composite Structures 183 (2018) 389–396
Fig. 11. Actual applied load-deflection responses of beams CSI-1 and CSI-2: a) at the
first cycle b) the 50th cycle c) at half of total cycles d) at pre-failure and failure.
CSI-1 resisted 114 cycles less than beam CSI-2 (only 0.4% differ-
ence), though, prestressing force was the same for both beams
(55 percent of ffu) (see Table 4). Regarding beams CSI-3 and CSI-
4, number of cycles varies more and the difference was 342 cycles
(about 4.5% difference) and N was 3.75 times less than the rest of
group 1 specimens. It is notable, that increasing the level of
prestress (15 percent) ultimate flexural strength is also increased
Fig. 12. Actual applied load-deflection responses of beams CSI-3 and CSI-4: a) at the
first cycle b) the 50th cycle c) at half of total cycles d) at pre-failure and failure.
(N greater approx. 4 times). Also, it shall be considered, that
increasing level of prestress, ffu is used more effectively. In regard
to beams CSI-5 and CSI-6, the same minimum load was kept
(20 kN), but the maximum load was 57 kN. Beam CSI-5 resisted
4336 cycles less than beam CSI-6. The difference consisted only 5
percent of ultimate number of resisted cycles (hereinafter-Nf).
These beams with standed 2.25 less of N than the rest of group 2
specimens (beams CSI-7 and CSI-8). Certainly, it shall be
 E. Atutis et al. / Composite Structures 183 (2018) 389–396
considered, that maximum load of beams CSI-5 and CSI-6 was 18
percent higher than remaining beams of the group 2. Also, beams
CSI-5 and CSI-6 resisted 3 times higher N than beams CSI-1 and
CSI-2 when the maximum load and prestressing force was 10 per-
cent less. Finally, results of group 3 specimens confirmed findings
from group 2 specimens e.g. decreasing maximum load by 18 per-
cent N of beams CSI-11 and CSI-12 increased 1.89 times. Consider-
ing same level of prestressing and minimum load, but 7 percent
less maximum load the final Nf values varied 2.4 times (compar-
ison of CSI-6 and CSI-10). Equal tendency was confirmed by com-
parison of beams CSI-8 and CSI-12, where at the similar
conditions Nf increased 2.02 times. Failure mode was governed
by rupture of the BFRP bars over the all beams.
5. Serviceability of BFRP prestressed concrete beams under
cyclic load
One of the aim of this research is to evaluate flexural stiffness
depending on increase of number of the load cycles. A comparison
between the cyclic load-deflection of group 1 specimens consider-
ing equal cyclic load amplitude is described further in this paper.
Applied cyclic load – deflection curves considering the first cycle,
the 50th cycle, a half of total cycles are presented in Figs. 11 and
12. The maximum deflection (hereinafter-dmax) of the beams CSI-
1 and CSI-2 at the first cycle was identical, correspondingly
5.0 mm and 5.6 mm as well as after application of the cyclic load-
ing within 50 cycles, 6.0 mm and 5.8 mm, respectively (see Fig. 11
(b)). The deflection of the beams CSI-3 and CSI-4 was higher than
the deflection at the first and 50th cycle at a given applied load,
8.6 mm and 10.9 mm, 9.2 mm and 12.8 mm, correspondingly (see
Fig. 12 (a) and (b)), though the level of prestress was 15 percent
less than beams CSI-1 and CSI-2. Once the number of cycles
reached a half of total cycles and after attaining the magnitude of
the cyclic load, the mentioned beams followed monotonic load-
deflection path. At this phase the deflection of beams CSI-1,
Fig. 13. Average crack width growth vs. fatigue life response a) group 1 specimens
b) group 2 specimens.
395
CSI-2, CSI-3 and CSI-4 was 10.8 mm, 10.6 mm, 12.4 mm and
16.6 mm, respectively. Interestingly enough, that beams pre-
stressed by BFRP jacked to 55 percent of ffu had less deflection than
beams prestressed by BFRP jacked to 40 percent of ffu, thus it con-
firm a greater stiffness due to degree of prestresing. Analysis of the
results of group 1 specimens shown, that dmax at pre-failure
(hereinafter-Nlim-1) was equal to the value at failure cycle (here-
inafter Nlim) (see Fig. 11 (d)). The deflection of the beams CSI-3
and CSI-4 at mentioned conditions was, correspondingly,
21.6 mm and 28.3 mm and was 35 to 86 percent higher than
beams CSI-1 and CSI-2. Again, it confirms, that a degree of prestress
shall to improve flexural stiffness of the beams and has a great
impact to final number of load cycles as well. The rest of the beams
(group 2 and group 3) of this research confirmed similar character
of applied cyclic load and deflection relationship, thus is not to be
presented in this paper additionally. Results of these specimens
were significant to predict a fatigue resistance as described previ-
ously in this paper.
Crack pattern and crack width (hereinafter-w) play significant
role in order to describe the behaviour of PC beams under cyclic
load. Further, it was observed an impact of a degree of prestress
and amplitude of cyclic load to crack width and growth with
respect to increase of the number of load cycles. Results of group
1 and 2 showed almost equal tendency of crack growth during
the experiment (see Fig. 13 (a) and (b)). Crack growth shall be
Fig. 14. Cracking pattern of a) group 1 b) group 2 c) group 3.
396 E. Atutis et al. / Composite Structures 183 (2018) 389–396
divided into 3 phases: until 20 percent of normalised N; from 20
percent to 70 percent of normalised N; and from 70 percent of total
N. Regarding phase 1, crack width increased more rapidly, whereas
at the phase 2 – can be described as more stable and, later, at phase
3, crack width increases again in most cases. It was confirmed by
the results of beams CSI-2, CSI-5, CSI-6 and CSI-7 (see Fig. 13).
The maximum average crack width (hereinafter-wmax) belongs
to beams CSI-3 and CSI-4 and were 0.64 mm and 0.62 mm respec-
tively. Also, the level of prestressed was 40 percent of ffu, whereas
the rest of analysed beams contained higher prestress (45 and 55
percent of ffu). It was observed, that crack width of beams CSI-1
and CSI-2 after application of cyclic loading within failure phase
and considering whole experiment time was identical to these of
group 2 specimens, though the prestress force was 10 percent
higher and maximum of cyclic load of beams CSI-5, CSI-6, CSI-7
and CSI-8 was 15 percent to 30 percent less than beams CSI-1
and CSI-2. Moreover, most of all beams performed with identical
crack pattern and failure crack occurred at applied concentrated
load place (at the end of pure bending zone) (see Fig. 14). Finally,
crack spacing and number of flexural cracks were identical taking
into account equal prestress level and cyclic load amplitude.
6. Conclusions
Cyclic loads over civil and industrial infrastructure pose an
urgent challenge for engineering community Composite materials
can play key role in meeting such challenges as a great potential
of fatigue resistance. Due to lack of complete understanding of fati-
gue strength of BFRP, experimental works were carried out in order
to develop a better understanding of the fatigue resistance as a
material (reinforcement) and structural element (reinforced con-
crete). The prestress was employed with the aim to use the high
strength composite material as much as more effective. Based on
above study, the major conclusions are as follows.
1) Firstly, BFRP bars shows sufficient resistance of the fatigue.
Further research shall be organised in order to predict more
accurate fatigue life of BFRP bars. The tests showed, that cyc-
lic stress range has a high influence to the fatigue life of BFRP
bars, whereas a frequency of the load varying from 2 Hz to
4 Hz does not extremely impact the results. Based on exper-
imental results a relationship of stress range and fatigue life
was proposed.
2) The level of prestressing has a significant impact to the fati-
gue strength of the beams. It was confirmed, that increasing
the level of prestress by 15 percent, the ultimate fatigue
strength is also increased by 3.8 times.
3) Stress caused by cyclic loading varying from 55 percent to 65
percent of BFRP bars tensile strength can be considered as a
safe range.
4) The tests showed that the number of total load cycles were
increased by 1.89 times when applied maximum load was
decreased by 10 percent for the equal (similar) beams. The
final number of load cycles increased more than 2 times
when maximum load was decreased 18 percent of beams
with the same level of the prestress and minimum applied
cyclic load.
5) In regard to serviceability, a degree of prestressing improves
flexural stiffness. Beams prestressed by BFRP jacked to 55
percent of the ultimate BFRP strength had less deflection
than beams prestressed by BFRP jacked to 40 percent of
the ultimate strength of BFRP. At a half of the total cycles
compared beams produces stabilization of the crack growth.
In particular, increasing prestress level by 15 percent, the
deflection at cracking stabilization (normalization) phase is
to be less 15 percent to 57 percent, and 35 percent to 86 per-
cent at failure phase of the beams. The maximum average
crack width occured in the beams at the least prestress level
jacked to 40 percent of the ultimate BFRP strength in com-
parison with beams prestressed by 55 percent of the ulti-
mate BFRP strength where the maximum average crack
width was 0.29 mm.
References
[1] Torres L, Sharaky I, Barris C, Maena M. Experimental study of the influence of
adhesive properties and bond length on the bond behaviour of NSM FRP bars in
concrete. J Civil Eng Manage 2016;22(6):808–17.
[2] Glišović I, Stevanović B, Petrović M. Bending behaviour of glulam beams
reinforced with carbon FRP plates. J Civil Eng Manage 2015;21(7):923–32.
[3] Skuturna T, Valivonis J. The statistical evaluation of design methods of the
load-carrying capacity of flexural reinforced concrete elements strengthened
with FRP. Arch Civil Mech Eng 2015;15(1):214–22.
[4] Marčiukaitis G, Valivonis J, Bareišis J. An analysis of the joint operation of a
CFRP concrete in flexural elements. Mech Comp Mater 2007;43(5):467–78.
[5] Skuturna T, Valivonis J. Experimental study on the effect of anchorage systems
on RC beams strengthened using FRP. Comp Part B: Eng 2016;91:283–90.
[6] Skuturna T, Valivonis J. Design method for calculating load-carrying capacity of
reinforced concrete beams strengthened with external FRP. Constr Build Mater
2014;50:577–83.
[7] Adimi M, Rahman A, Benmokrane B. New method for testing fiber–reinforced
polymer rods under fatigue. J Comp Constr 2000;4(4):206–13.
[8] Noël M, Soudki K. Fatigue behavior of GFRP reinforcing bars in air and in
concrete. J Comp Constr 2014;18(5):1–8.
[9] Terrasi G, Affolter Ch, Barbezat M. Numerical optimization of a compact
reusable pretensioning anchorage system for CFRP tendons. J Comp Constr
2011;15(2):126–35.
[10] El-Refai A. Durability and fatigue of basalt fiber reinforced polymer bars
gripped with steel wedge anchors. J Comp Constr 2013;17(6):1–11.
[11] Noël M, Soudki K. Fatigue behavior of full-scale slab bridge strips with FRP
reinforcement. J Comp Constr 2015;19(2):1–9.
[12] El-Refai A, Jeffrey S, Soudki K. Performance of CFRP tendon-anchor assembly
under fatigue loading. Comp Struct 2007;80(3):352–60.
[13] Sendeckyj GP. Fitting models to composite materials fatigue data. Test
methods and design allowables for fibrous composites. In: ASTM STP
734. American Society for testing and Materials; 1981. p. 245–60.
[14] El-Ragaby A, Salakawy E, Benmokrane B. Fatigue life evaluation of concrete
bridge deck slabs reinforced with glass FRP composite bars. J Comp Constr
2007;11(3):258–68.
[15] Wu Zh, Wang X, Iwashita K, Sasaki T, Hamaguchi Y. Tensile fatigue behavior of
FRP and hybrid FRP sheets. Comp Part B: Eng 2010;41(5):396–402.
[16] Brondsted P, Lilholt H, Anderson SI. Fatigue damage prediction by
measurements of the stiffness degradation in polymer matrix composites.
Proceedings of the International Conference on Fatigue Composites Eight
International Spring Meeting. p. 370–7.
[17] Rezansoff T, Zacaruk JA, Afseth JG. High cycle (fatigue) resistance of reinforced
concrete beams with lap splices. Can J Civil Eng 1993:642–9.
[18] Talreja R. Fatigue of composite materials. Damage mechanisms and fatigue-life
diagrams. Proc. R. Soc. London. Ser. A, Math. Phys. Sci. 1981;378:461–75.
[19] fib bulletin 40. FRP reinforcement in RC structures. Lausanne (EPFL):
International Federation for Structural Concrete; 2007.
[20] ACI 440.3R-12. Guide test methods for fiber-reinforced polymer (FRP)
composites for reinforcing or strengthening concrete and masonry
structures. Farmington Hills (MI): American Concrete Institute; 2012.
[21] ACI 440.4R-04. Guidelines for prestressing concrete structures with FRP
tendons. Farmington Hills (MI): American Concrete Institute; 2004.
[22] ISO 10406-1:2008 (E). Fibre-reinforced polymer (FRP) reinforcement of
concrete – Test methods, Part 1: FRP bars and grids, 2008.
[23] Design Manual No. 5. Prestressing concrete structures with fibre-reinforced
polymers. Winnipeg (MT): ISIS Canada; 2008.