Engineering Structures 206 (2020) 110088

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Shear strength of prestressed FRP reinforced concrete beams with shear T

reinforcement
⁎
Sophia Kueres , Norbert Will, Josef Hegger
RWTH Aachen University, Institute of Structural Concrete, Mies-van-der-Rohe-Str. 1, 52074 Aachen, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: To avoid brittle shear failures, the shear strength of reinforced concrete beams has to be verified at ultimate limit
CFRP state. For this purpose, simple but accurate shear design provisions or shear resistance models are required. In
Prestressing current design provisions the influence of prestressing is either neglected or insufficiently considered in the
Shear tests design equations leading to rather conservative results due to the lack of investigations concerning the shear
Shear resistance model
strength of FRP prestressed beams. To overcome this drawback, shear tests on concrete beams with textile shear
Variable strut inclination
reinforcement and pre-tensioned CFRP tendons were conducted and the results are presented in this paper.
Besides the investigation of the influence of the level of prestressing and of the shear span-depth ratio on the
shear strength, special attention was paid to the formation and the inclination of the shear cracks as well as the
crack pattern to investigate the failure mechanism more in detail. Furthermore, theoretical investigations were
conducted in order to derive a simple shear design model for prestressed concrete beams with FRP shear re-
inforcement. For this purpose, an existing shear resistance model based on a truss model with variable strut
inclination is extended considering the beneficial effects of prestressing on shear strength.

1. Introduction model is necessary.

To achieve this objective, in a first step, a variable strut inclination
In general, the behavior of beams with FRP reinforcement under model was derived to determine the shear strength of non-prestressed
shear loading is similar to the behavior of steel reinforced concrete FRP reinforced beams with shear reinforcement and validated by means
beams [1]. Nevertheless, the use of FRP as shear and longitudinal re- of databank evaluations [11]. In this paper, an approach is proposed to
inforcements affects the shear capacity due to the different mechanical extend the developed model to take into account the beneficial effects
properties, such as the lower modulus of elasticity and the linear-elastic of prestressing on shear strength. Since the available databank to va-
behavior until failure. To investigate the aforementioned differences lidate the extended shear resistance model is very limited, additional
more in detail, extensive research on the shear behavior of FRP re- tests were carried out. In this context, nine shear tests on beams with
inforced concrete beams with shear reinforcement [2–5] has been textile shear reinforcement and pre-tensioned CFRP tendons were
conducted over the last years. Based on the results, various guidelines conducted and the results are presented. Besides the investigation of the
have been developed [1,6–8]. influence of the level of prestressing and of the shear span-depth ratio
In contrast to the numerous test series, which were conducted to on the shear strength, the formation and the inclination of the shear
investigate the shear behavior of non-prestressed beams with FRP shear cracks as well as the crack pattern is analyzed based on the test results
reinforcement, the number of tests series evaluating the beneficial in- to investigate the failure mechanism more in detail.
fluence of prestressing on the shear strength of FRP reinforced beams is
rather limited [9,10]. Due to this lack of investigations concerning the 2. Experimental investigations
shear strength of FRP prestressed beams, the influence of the prestres-
sing is either neglected or insufficiently considered in the existing de- 2.1. General
sign formulas leading to rather conservative results [10]. To overcome
this drawback and to allow for an economical and safe shear design of The shear strength of concrete beams with pre-tensioned CFRP
both non-prestressed and prestressed beams with FRP shear reinforce- tendons and textile CFRP shear reinforcement was investigated on
ment, the development of a simple, generally applicable shear design simply supported T-beams. The main objective of these tests was the

⁎
Corresponding author.
E-mail addresses: [email protected] (S. Kueres), [email protected] (N. Will), [email protected] (J. Hegger).

https://doi.org/10.1016/j.engstruct.2019.110088
Received 26 June 2019; Received in revised form 6 November 2019; Accepted 11 December 2019
0141-0296/ © 2019 Elsevier Ltd. All rights reserved.
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Notation El elastic modulus of longitudinal reinforcement

Ep elastic modulus of the prestressing reinforcement
Roman lower case letters Et elastic modulus of the longitudinal textile reinforcement
Ew elastic modulus of shear reinforcement
a shear span Fmax failure load
a/d shear span-depth ratio M sectional moment
aw cross-sectional area of the shear reinforcement per unit N sectional normal force
length V sectional shear force
b width of compression zone Vcalc calculated shear capacity
bw web width Vcs shear capacity of compression strut
d effective depth VR shear resistance
dg maximum aggregate size Vs shear contribution of shear reinforcement
fce effective concrete compressive strength Vtest experimental shear strength
fcm mean value of concrete compressive strength
ffm mean tensile strength of FRP reinforcement Greek letters
fpj jacking stress of tendon
ftm mean tensile strength of textile reinforcement Δe eccentricity of normal force
fw tensile strength of shear reinforcement Δh change in thickness
s spacing of shear reinforcement ηfc strength reduction factor considering the brittle fracture
z inner lever arm behavior of concrete with higher compressive strength
zl inner lever arm of longitudinal reinforcement ε1 principal tensile strain
zp inner lever arm of prestressing reinforcement ε2 principal compressive strain
zs vertical center of gravity of the beam εx horizontal strain at mid-depth of compression field
zt inner lever arm of longitudinal textile reinforcement θ inclination of compression strut
v strength reduction factor considering strain effects
Roman upper case letters ρw shear reinforcement ratio
ρw* related shear reinforcement ratio
Al cross-sectional area of the longitudinal reinforcement σcp concrete compressive stress due to prestressing
Ap cross-sectional area of the prestressing reinforcement
At cross-sectional area of the longitudinal textile reinforce- Other symbols
ment
Aw cross-sectional area of the shear reinforcement # number of tests
Ec elastic modulus of the concrete

determination of the influence of the level of pre-tensioning and of the but different tread depths were applied. The geometric and material
shear span-depth ratio on the shear strength as well as the investigation properties of both tendon types derived from tests are given in Table 2.
of the shear crack formation and crack geometry. Additional information concerning the bond properties of the tendons
are reported in [13] and [14], where experimental investigations on the
bond behavior of the CFRP tendons were conducted.
2.2. Materials
Due to the high capacity of the CFRP reinforcements in tension, a
high strength concrete (HSC) was selected for the fabrication of the test
In all beam specimens of the test series presented in this paper,
specimens to provide a sufficient concrete compressive strength and to
textile reinforcement was applied as shear reinforcement and additional
prevent premature failure of the concrete and, thus, the specimens. To
longitudinal reinforcement. The textile reinforcement is fabricated as
ensure a simple casting and compacting of test specimens reinforced
carbon filaments bundled in rovings and impregnated with epoxy resin.
with textile reinforcement and pressure-sensitive CFRP tendons, a self-
The individual rovings are arranged crosswise with an axial spacing of
compacting concrete with a limited maximum aggregate size was re-
38 mm. The main geometric and material properties of the textile CFRP
quired. Based on this specifications, a fine-grained HSC (dg = 5 mm)
reinforcement given by the manufacturer and derived from tests in [12]
with a strength class C70/85 was selected, which has been used in
are summarized in Table 1. The stress-strain curve of the textile re-
previous studies [15]. The main properties of the concrete at test days
inforcement is linear-elastic.
are provided in Table 3. In this context it should be noted, that the
As tendons, CFRP bars with a diameter of 10 mm and a profiled
concrete strength at test days was higher than initially expected.
surface were used in the experimental investigations. The profiled
surface is generated by a spirally wrapping with impregnated CFRP
fibers to enhance the bond properties of the bars. In the tests specimens
two different tendon types T1 and T2 with identical material properties

Table 1
Geometric and material properties of textile CFRP reinforcement given by manufacturer (and by [12]).
Orientation at Aroving ftm Et εtu
[–] [mm2/m] [mm2] [MPa] [MPa] [‰]

0° 142 5.42 2500 (2700) > 180,000 (205,000) 14.0 (13.2)

90° 142 5.42 2800 (2900) > 180,000 (230,000) 15.5 (12.4)

at: cross-sectional area of the textile reinforcement; Aroving: cross-sectional area of one roving; ftm: mean value of the tensile strength of the textile reinforcement; Et:
elastic modulus of the textile reinforcement; εtu: ultimate tensile strain of the textile reinforcement.

2
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Table 2 2.4. Experimental results

Geometric and material properties of CFRP tendons.
Tendon Tread depth Fiber volume ratio Ap fpm Ep εpu 2.4.1. General
[–] [mm] [%] [mm2] [MPa] [MPa] [‰] In Table 4, the main test results, such as cracking and failure loads,
related shear capacities, range of measured crack angles of all shear
T1/T2 0.5/0.05 53.0 41.6 3610 187,000 19.3
cracks, as well as failure types are summarized. In all tests on specimens
Ap: fiber cross-sectional area of the tendon; fpm: mean value of the tensile
LV-1 and LV-2, a bond slip failure of the tendons (T2) was observed
strength of the tendon; Ep: elastic modulus of the tendon; εpu: ultimate tensile prior to shear failure. Therefore, in these tests the overhang of the beam
strain of the tendon. at the support was increased from test to test to prevent this failure type
Unfortunately, the increased overhang was not sufficient to prevent this
2.3. Test setup and procedure failure mode. The bond failure can be explained by the poor bond
quality of the tendon T2, caused by the lower tread depth compared to
In total, nine shear tests were carried out on seven beams with two T1 (see Table 2), resulting in an insufficient anchorage of the tendons.
different cross sections. Both cross sections were reinforced with four Furthermore, test specimen LV-4 failed by bond slip of a bottom tendon
CFRP tendons (T1 or T2) and a textile shear reinforcement was applied (T1) caused by a splitting crack at the unloaded end of the beam, which
over the entire beam length. Six beams were pre-tensioned with a occurred at release of the pre-tensioning stress. The remaining four
jacking stress fpj between 1460 and 1820 MPa, which was released after tests, prestressed with T1 tendons, showed a shear induced failure by
one day of concrete hardening. Also, one reference beam without pre- rupture of the textile shear reinforcement.
stressing force was tested. Once the concrete reached a sufficient
compressive strength (~fcm,cyl,28 = 92 MPa), shear tests were carried 2.4.2. Influence of shear slenderness
out on single span beams subjected to point loads to determine the Fig. 2 presents the normalized shear force-deflection curves sepa-
shear capacities and the related failure modes. The test setup of the rately for the different cross sections and the tested influencing para-
shear tests is shown in Fig. 1(a, b). The two cross sections of the beams meters. Even though a bond failure occurred in all tests on beams LV-1
are illustrated in Fig. 1(c, d). In total, five beams with the cross section and −2, a comparison of the development of the load-deflection curves
shown in Fig. 1(c) and a length of 2.65 m were tested. In addition, shear (Fig. 2(a)), indicates an influence of the shear span-depth ratio a/d on
tests on two beams with a larger cross section (Fig. 1(d)) were con- the shear strength. With increasing ratio a/d the shear resistance de-
ducted. The length of these beams was increased to 5.0 m to allow for creases. This observation is confirmed by the results of tests LV-3 and
two tests on one beam. To investigate the effect of shear slenderness on LV-7 (Fig. 2(b)). However, the influence is significantly more pro-
the shear strength, the distance between load application point and nounced for the smaller test specimens. A comparison of the normalized
support was varied for some test specimens. The main geometric shear strengths reveals similar values for the test specimens of both
parameters and material properties of all beams are summarized in beam types with a shear slenderness of 3.5 indicating no influence of
Table 3. the effective depth. For the test specimens with a/d = 2.5 the smaller
During the test, the test load and deflection at load introduction as beam reached even a higher value than the larger beam.
well as the concrete compression strain and vertical and diagonal dis-
placements on the concrete surface were measured continuously using 2.4.3. Influence of prestressing
strain gages or linear variable displacement transducers (LVDTs). Besides the influence of the shear slenderness, the effect of the level
Furthermore, the end slip of the tendons was monitored via LVDT of pre-tensioning was investigated in test specimens LV-3 to LV-6. The
measurements to identify bond slip failure at the loaded end of the jacking stress fpj of the beams was varied between 1580 and 1820 MPa
beam. The load was applied in steps of 1/8 of the estimated failure load (44 to 50% of fpm) and an additional reference beam without pre-ten-
Fmax. Until the first flexural crack, the load was applied force controlled sioning stress was tested. Unfortunately, a splitting crack occurred
with 3.5 min/step. Subsequently, the type of loading was changed to during the release of the jacking stress of test specimen LV-4
displacement control with a speed of 0.5 mm/min, which corresponds (fpj = 1820 MPa), which led to a premature failure caused by bond slip
approximately to the deflection rate measured during the first test of the tendon at the unloaded beam end. The shear force–deflection
phase. Hence, a constant load speed in both the uncracked and cracked curves of the test series are shown in Fig. 2(c). As expected, the shear
state of the beam was attained. strength increases with increasing pre-tensioning stress. The compar-
ison of the load-deflections curves of the pre-tensioned beams with the
reference beam demonstrates the significantly higher cracking loads
caused by the prestressing. After the first flexural cracks occurred, the
load-deflection curves of all beams are more or less parallel until failure

Table 3
Main parameters of shear tests.
Specimen Tendon l dm a1/a2 e1/e2 a/d ρw fpj fcm,cyl,test Ecm,test
[–] [–] [mm] [mm] [mm] [mm] [–] [%] [MPa] [MPa] [MPa]

LV-1-1 T2.2 5000 350 875/3925 100/100 2.5 0.19 1460 102.0 44,200
LV-1-2 T2.2 5000 350 875/3775 250/100 3.5 0.19 1460 112.9 45,200
LV-2-1 T2.2 5000 350 1225/2475 300/1000 2.5 0.19 1460 113.0 45,100
LV-2-2 T2.2 5000 350 1225/1925 500/1350 3.5 0.19 1460 115.1 45,000
LV-3 T2.1 2650 226 790/1360 400/100 3.5 0.19 1580 99.2 45,500
LV-4 T2.1 2650 226 790/1360 400/100 3.5 0.19 1820 95.2 42,500
LV-5 T2.1 2650 226 790/1360 400/100 3.5 0.19 – 86.2 44,800
LV-6 T2.1 2650 226 790/1360 400/100 3.5 0.19 1670 92.1 41,600
LV-7 T2.1 2650 226 565/1585 400/100 2.5 0.19 1580 85.2 40,900

l: beam length; dm: average effective depth of the tendons; a1/a2: shear span; e1/e2: overhang at support; a/d: governing shear slenderness; ρw: shear reinforcement
ratio; fpj: jacking stress of tendons; fcm,cyl,test: mean value of the concrete compressive strength at day of testing; Ecm,test: elastic modulus of the concrete at day of
testing.

3
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Fig. 1. Test setup of shear tests (a), (b), cross section of test specimens LV-3 – 7 (c), and cross section of test specimens LV-1 – 2 (d).

Table 4 Table 5
Results of shear tests. Measuring lengths and angles of the LVDTs in shear tests.
Specimen w Ffl,cr Fs,cr Fmax V β Failure Specimen H1/H2 V1/V2 D1/D2 α
[–] [mm] [kN] [kN] [kN] [kN] [°] [–] [–] [mm] [mm] [mm] [°]

LV-1-1 17.9 109 210 227 186 40–55 B LV-1 – 2 280 335 355 50
LV-1-2 22.4 98 130 267 179 30–48 B LV-3 – 7 280 225 250 37
LV-2-1 31.4 115 150 258 209 28–47 B
LV-2-2 44.9 110 210 236 138 47–54 B α: measuring angle of LVDTs D1/D2; H1,H2,D1,D2,V1,V2: LVDTs (notation refers
LV-3 22.7 95 147 281 178 31–46 RS to Fig. 1).
LV-4 18.6 100 200 279 177 31–40 B
LV-5 33.0 20 130 231 146 33–43 RS
LV-6 26.5 95 230 293 185 29–40 RS LV-2 showed no shear induced failure, their results are of limited sig-
LV-7 18.2 105 140 330 243 33–43 RS nificance and are not discussed in this paper.
In Fig. 3(a, b), the displacement measurements of the vertical LVDTs
w: maximum deflection at load introduction; Ffl,cr: flexural cracking load ; Fs,cr: are presented for all beams. Due to the shorter distance between load
shear cracking load; Fmax: maximum failure load; V: shear force (without self introduction and support for test specimen LV-7 (a/d = 2.5), the po-
weight); β: crack angle; B: bond slip failure; RS: rupture of shear reinforcement.
sition of LVDT V1 would have been directly over the support and
therefore was omitted. As expected, the vertical displacements V1 closer
indicating a similar structural behavior. to the support are clearly smaller than the displacements V2 since the
majority of cracks is located in the vicinity of the load introduction (see
2.4.4. Cracking characteristics Fig. 5(a)). Nevertheless, the governing shear failure crack often devel-
To enable a more detailed analysis of the structural behavior of the oped closer to the support leading to vertical displacements V1 of about
beams under shear loading, the horizontal, vertical, and diagonal dis- 2 to 4 mm (Fig. 3(a)). The small deformations (0.7 to 0.9 mm) measured
placements of the concrete surface in the shear span were measured for test specimen LV-4 verify the premature bond slip failure of the
with LVDTs on one side of the web. The measuring lengths and angles tendon.
of the applied LVDTs are given in Table 5. The results of the analysis of Fig. 4 shows the displacement measurements of the diagonal LVDTs
the crack formation and geometry are presented in the following sec- of the beams. A comparison of Fig. 4(a) and (b) reveals a similar result
tion for test specimens LV-3 to LV-7. Since the tests on beams LV-1 and as observed for the vertical displacements in Fig. 3. The displacements

Fig. 2. Comparison of normalized shear force-deflection curves: Influence of shear span-depth ratio (a, b), and influence of level of pre-tensioning (c).

4
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Fig. 3. Vertical displacement measurements V1 (a) and V2 (b) of test specimens LV-3 to LV-7.

D2 measured near the load introduction are significantly larger than the other researchers to validate the own conclusions. However, only two
displacements D1. Moreover, the small values of D1 measured in all tests studies investigating the shear strength of beams pre-tensioned with
(< 1.0 mm) indicate a mostly vertical crack opening of the governing CFRP tendons and with FRP shear reinforcement, presented in [9] and
failure crack. The displacements D2 vary between 3 and 5 mm for all [10], are available in literature. This proves the very limited database of
beams and show a smaller variation compared to the vertical dis- shear tests on pre-tensioned beams with FRP shear reinforcement and,
placements V2 (Fig. 3(b)). To further investigate the crack formation thus, the lack of basis for the derivation of valid resistance and design
during the test, the changes in thickness of the beams for different load models. Therefore, in addition to the four beams of the own tests, which
levels are presented in Fig. 5(b). Furthermore, the crack pattern before failed in shear, two of the four tests on larger beams, which failed due to
and after failure is illustrated in Fig. 5(a) and (c) to enable a profound bond slip, were taken into account for comparison reasons. In this
analysis of the measured values. context, the maximum load attained for both shear span-depth ratios
Fig. 5(a) reveals a very similar crack pattern prior to failure of was considered as a conservative estimation of the actual shear strength
specimens LV-3 and LV-4 with several shear cracks, which are evenly of the specimens. In Table 6, the main parameters of the considered
distributed between the LVDTs V1 and V2. In contrast, the crack pattern shear tests are summarized.
of LV-6 shows only one shear crack at the position of V1 and several All beams were pre-tensioned with CFRP tendons resulting in con-
flexural cracks closer to the load introduction. This difference could crete compressive stresses in the range of -σcp = 3.0–5.3 MPa. The shear
also be observed by the analysis of the measured changes in thickness of reinforcement used by [9] was also a textile CFRP reinforcement pro-
the web Δh presented in Fig. 5(b). For LV-3 and LV-4, Δh is approxi- duced by the same manufacturer as the textile reinforcement used for
mately equal at both positions, whereas a significantly higher de- the own tests. In the tests conducted by [10], GFRP stirrups were used
formation in the area of the flexural cracks is monitored for LV-6. In as shear reinforcement. The main parameters investigated in the test
general, the values reported for LV-4 are considerably smaller series were the shear span-depth ratio a/d and the compressive stress
(< 1.0 mm) due to the premature bond failure. The comparison of the -σcp. In this context, Fig. 6 shows the normalized shear strengths Vtest/
changes in thickness of test specimens LV-3 and LV-6 indicates an in- (bw·√fcm) depending on the influencing parameters. In Fig. 6(a), a clear
itiation of the shear crack formation between 70 and 80% of the ulti- trend regarding the effect of the shear slenderness can be observed for
mate failure load. The largest increase of Δh, about 60 to 80% of the all tests. An increasing ratio of a/d correlates with a decreasing shear
total deformation, is measured between 95% of the failure load and strength and the effect is more pronounced for smaller effective depths.
failure. In Fig. 5(c), the governing shear crack of test specimens LV-3 These conclusions correspond to the observations made in the own tests
and LV-6 can be identified as the closest shear crack to the support with (see Fig. 2(a, b)). For comparison reasons the influence of the shear
a crack angle of about 30°. The measured shear crack angles of the other slenderness according to EC2 [16] up to a ratio of a/d = 2.0 is illu-
test specimens are in the same range (see Table 4). Based on the eva- strated. Moreover, the adapted influence for further limit values of a/
luation, the failure type can be determined as rupture of the textile d = 2.5 and 3.0 is depicted. Considering all test results, an influence of
shear reinforcement at the bent portion of the reinforcement cage. The the shear slenderness up to a value of a/d = 3.0 seems to be verified.
extensive spalling of the concrete cover between two shear cracks can Nevertheless, to prove this assumption, further tests with a higher shear
be classified as a secondary failure. slenderness (a/d > 4.0) have to be conducted in the future. In
Fig. 6(b), the influence of the compressive stress due to pre-tensioning
on the shear strength is depicted. Since in [10] no reference tests
2.4.5. Comparison with test results from literature without prestressing were performed, the tests of this study are not
In the following, the own test results are compared to the results of

Fig. 4. Diagonal displacement measurements D1 (a) and D2 (b) of test specimens LV-3 to LV-7.

5
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Fig. 5. Crack pattern prior to failure (a), measured changes in thickness for different load levels and locations (b), and failure crack pattern (c) of test specimens LV-3,
LV-4, and LV-5.

considered for the comparison. investigations presented in Section 2, was implemented in the model.
As expected, all test showed an increase of the shear strength with
increasing compressive stress. However, the effect is significantly more 3.2. Variable strut inclination model for shear design of FRP reinforced
pronounced for the test specimens tested by [9] than for the own tests. concrete members with shear reinforcement
A possible explanation might be the different cross sections and pre-
tensioning conditions of the investigated beams. The structural beha- According to the variable strut inclination model proposed in [11]
vior of the T-beams tested in the own study, is rather similar to a rec- the shear strength of a beam with shear reinforcement Vs can be de-
tangular cross section due to the large web width (see Fig. 1(b, c)). In termined by:
contrast, in [9] considerably slender I-beams with a web width of only
40 mm were tested. Moreover, besides the pre-tensioned tendons in the Aw
Vs = ·f ·z·cotθ
bottom flange an additional prestressing tendon was applied to the s w (1)
upper flange of these beams. A similar load increasing effect due to the where Aw is the cross-sectional area of the shear reinforcement, s is the
flanges of prestressed I-beams has already been identified for steel re- spacing of the shear reinforcement (stirrups), fw is the effective tensile
inforced concrete beams as reported in [17] or [18]. strength of the FRP shear reinforcement, z is the inner lever arm, which
may be assumed as 0.9·d, and θ is the angle of the compression strut.
3. Theoretical investigations The capacity of the truss is limited by the compressive stress fce in the
inclined concrete struts, which is depending on their inclination θ.
3.1. General Hence, the maximum capacity of the strut Vcs is defined as:
fce ·bw ·z
In the following section, theoretical investigations are presented in Vcs =
tanθ + cotθ (2)
order to derive a simple shear design model for both prestressed and
non-prestressed beams with FRP shear reinforcement. In this context, Consequently, the governing shear strength VR is defined as the
the shear resistance model proposed in [11], which is based on a truss minimum value of Vs or Vcs. In contrast to the shear design of steel
model with variable strut inclination, is extended considering the reinforced beams, for the design of FRP reinforced beams, the angle of
beneficial effects of prestressing on shear strength. Furthermore, the the compression strut θ is not variable, but dependent on the re-
capability to consider shaped textile reinforcement as shear reinforce- inforcement ratios, material properties and loading conditions [11]. A
ment, such as the shear reinforcement applied in the experimental suitable solution to determine the angle θ is the application of the

Table 6
Shear tests on beams with FRP shear reinforcement and pre-tensioned with CFRP tendons as well as non-prestressed reference tests.
Researcher # d a/d ρl El ρw Ew fw fcm -σcp
[–] [–] [mm] [–] [%] [MPa] [%] [MPa] [MPa] [MPa] [MPa]

Kromoser [9] 4 230 3.0–4.0 0.57 161,000 0.23 – 1300 136 0–3.8
Kueres 2018 6 226–350 2.5–3.5 0.50 187,000 0.19 180,000 1960 85–99 0–5.3
Nabipay [10] 6 189 1.5–3.5 1.08 136,000 0.20 40,800 827 48 3.0–3.3

#: number of tests; d: effective depth; a/d: shear span-depth ratio; ρl: longitudinal reinforcement ratio; El: elastic modulus of longitudinal reinforcement; ρw: shear
reinforcement ratio; Ew: elastic modulus of shear reinforcement; fw: tensile strength of the shear reinforcement determined in test; fcm: mean value of the concrete
compressive strength; -σcp: concrete compressive stress due to pre-tensioning.

6
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Fig. 6. Influence of the shear span-depth ratio (a) and the compressive stress (b) on the shear capacity of pre-tensioned concrete beams.

compression field model (CFM) originally developed by SIGRIST [19,20] ν=

1
for steel reinforced concrete beams. Based on the findings of SIGRIST, in 1.2 + 55·ε1 (7)
[11], the following equation to determine the inclination of the com-
The principal strain ε1 is defined by Eq. (8) based on Mohr’s circle of
pression struts of the compression field was derived:
strains:
1/3
V ε1 = εx + (εx − ε2)·cot 2θ (8)
cotθ = ⎛ ⎜ ⎟
⎞
⎝ εx · z · a w · E w ⎠ (3)
where εx is given by Eq. (4) and the principal strain ε2 is assumed as
where V is the sectional shear force, εx is the horizontal strain of the −0.002 for all concrete types [20].
compression field at mid-depth, aw = Aw/s is the cross-sectional area of
the shear reinforcement per unit length and Ew is the elastic modulus of 3.3. Application of variable strut inclination model for prestressed beams
the shear reinforcement. Considering the compression field illustrated and beams with textile shear reinforcement
in Fig. 7, the horizontal strain of the compression field εx may be cal-
culated using the simplified approach given by Eq. (4). In this context, Since the databank evaluation presented in [11] proved the suit-
the strain in the compression chord is neglected as a conservative as- ability of the proposed resistance model to predict the shear strength of
sumption. non-prestressed beams with FRP shear reinforcement (mean value
M
+ V + 0.5·N μx = 1.01, coefficient of variation Vx = 0.18), in a next step, the ex-
z
εx = tension of the model to take into account the influence of prestressing
2·El·Al (4) was possible. Moreover, the capability to consider other types of shear
In Eq. (4), M, V, and N are the applied sectional forces in the de- reinforcement than stirrups, e.g. shaped textile reinforcement, was
cisive section for the verification of the shear strength and El and Al are implemented. In this context, Fig. 8 shows a prestressed concrete beam
the elastic modulus and the cross-sectional area of the longitudinal with FRP prestressing tendons, a textile shear reinforcement, and FRP
reinforcement. Since both εx and cotθ depend on the applied shear force bars as flexural reinforcement. When a textile shear reinforcement is
V, the shear strength has to be determined iteratively. The calculation applied, automatically an additional longitudinal textile reinforcement
procedure is explained in detail in [11]. is provided due to the crosswise structure of the textile reinforcement.
Besides the angle of the compression struts θ, the tensile strength of This effect has to be considered in the determination of the inner lever
the shear reinforcement fw and the effective compressive strength of the arm z and of the strain at mid-depth εx.
compression struts fce are required to determine the governing shear For members containing various types of longitudinal reinforce-
strength of a cross section according to the resistance model (see Eqs. ment, the effective lever arm z can be calculated as follows:
(1) and (2)). The tensile strength of the shear reinforcement fw depends z l2·Al + z p2·Ap + zt2·At
on the ultimate tensile strength of the FRP reinforcement ffm and the z=
z l ·Al + z p·Ap + z t ·At (9)
type of shear reinforcement. If the bars are bent during the production
of shear reinforcement, like for FRP stirrups, a reduced FRP tensile where zl and Al are the lever arm and cross-sectional area of the long-
strength has to be applied, which should be provided by the manu- itudinal reinforcement, zp and Ap are the values of the prestressing re-
facturer based on tensile tests. The evaluation of the material properties inforcement and zt and At are the values of the longitudinal textile re-
of the stirrups applied as shear reinforcement in the shear tests of the inforcement. Regarding the textile reinforcement, only the
databank presented in [11] revealed an average value of fw that cor- reinforcement located in the tension zone of the cross section should be
responds to 60% of ffm. considered as illustrated in Fig. 8. As a conservative estimation, the
The effective concrete compressive strength fce is defined according height of the tension zone can be assumed as d – zs, where d is the
to Eq. (5) following the approach of [20]: effective depth of the reinforcement and zs is the center of gravity of the
fce = ηfc ·ν ·fcm (5)
where ηfc is a strength reduction factor considering the brittle fracture
behavior of concrete with higher compressive strength and ν is a re-
duction factor accounting for strain influences. The strength reduction
factors are defined according to Eqs. (6) and (7) following the approach
presented in [20], which is based on investigations of RUIZ AND MUTTONI
[21] and KAUFMANN [22].
1/3
30
ηfc = ⎜⎛ ⎟⎞ Fig. 7. Determination of longitudinal strain at mid-depth of the compression
⎝ fcm ⎠ (6) field (based on [20]).

7
S. Kueres, et al. Engineering Structures 206 (2020) 110088

Fig. 8. Determination of longitudinal strain at mid-depth of web for members with textile shear reinforcement and/or prestressed tendons.

beam. Subsequently, the horizontal strain at mid-depth εx of a pre- strength of textile shear reinforcement should be investigated in the
stressed cross section with textile reinforcement can be determined future in appropriate tests similar to the tests for FRP bent bars and
using Eq. (10). stirrups.
In Fig. 9, the results of the evaluation of the extended shear re-

εx =
M
z
+ V + N· ( 1
2
±
Δe
z ) sistance model are presented. In this context, the ratio Vtest/Vcalc is il-
lustrated as a function of the related shear reinforcement ratio ρw* =
zp
2· ( zl
z
·El·Al + z
·Ep·Ap +
zt
z
·Et ·At ) (10) ρw·Ew/Ec. Fig. 9(a) shows the comparison of the calculated and the ex-
perimental shear strength of beams with textile shear reinforcement.
where Δe is the eccentricity of the acting normal force with respect to Moreover, the influence of the additional longitudinal textile re-
the neutral axis of the cross section. If the value of εx is negative it shall inforcement At on the shear strength is illustrated. The “shear strength
be taken as zero. The internal forces caused by the prestressing have to without At“ is calculated with εx according to Eq. (4), whereas the
be considered in the sectional forces M, V and N. For a prestressed beam “shear strength with At“ is determined applying Eq. (10). In this con-
with stirrups instead of textile shear reinforcement or a non-prestressed text, the shear strengths of the tests carried out by [24] are not sig-
beam with textile shear reinforcement, the respective terms are ne- nificantly affected by the additional At as the influence is rather small
glected. due to the low section height and a single-legged shear reinforcement.
To evaluate the extended shear resistance model, existing data of Furthermore, the experimental shear strength of these tests is under-
shear tests on beams with textile shear reinforcement and shear tests on estimated by 20% resulting in a ratio of Vtest/Vcalc of approximately
prestressed beams were used. Unfortunately, the database of both beam 1.25. This can be explained by the use of fiber reinforced concrete,
types is very limited. The main parameters of the shear tests on pre- which was applied to prevent a splitting failure of the beams. Never-
stressed beams were already presented in Table 6. In this context, only theless, the additional capacity provided by the fibers is not considered
beams with a/d ≥ 3.0 were considered since for lower values a strong in the model.
influence of the shear slenderness on the shear strength was observed. The shear strengths of the own test and of the tests of [9] with a web
The main parameters of the existing shear tests on non-prestressed width of 40 mm (ρw* ≈ 0.80%), show a higher influence of the addi-
beams with textile shear reinforcement are summarized in Table 7. tional longitudinal reinforcement of the textile reinforcement. The
The textile CFRP reinforcement used in all tests is produced by the shear strengths calculated without At underestimate the experimental
same manufacturer. The elastic modulus of the shear reinforcement strength by 10–15%, whereas the shear strengths calculated with At
used in [9] is not known and, thus, conservatively assumed to be indicate a good correlation with the experimental results. The further
equivalent to the modulus of the reinforcement used in own tests tests conducted by [9] with a web width of 30 mm failed by a dela-
(Ew = 180,000 MPa). Due to the construction of the shear reinforce- mination of the concrete in the web due to the splitting effect of the
ment, the tensile strength fw of the shear reinforcement used in [9] and textile shear reinforcement. Consequently, the shear capacity of these
[24] corresponds to the mean tensile strength ftm and has not to be tests is overestimated by the proposed resistance model.
reduced. Regarding the own tests, an appropriate reduction factor for In Fig. 9(b), the same influence of the additional longitudinal re-
the tensile strength of bent textile shear reinforcement had to be spe- inforcement At is observed for the beams pre-tensioned with CFRP
cified since experimental investigations on this topic had not yet been tendons, if applicable. Considering At, the predicted shear capacities of
conducted. Due to the different shape of the cross section of rovings the own tests indicate a good correlation with the experimentally de-
compared to stirrups, rather elliptical than circular, and the con- termined shear strengths resulting in a mean value of Vtest/Vcalc = 1.05.
siderably smaller diameter, a lower influence of the bent portion on the In contrast, the shear resistance of the tests by [9] and [10] is under-
tensile strength of textile shear reinforcement was assumed. In this estimated by up to 25% according to the proposed model. This differ-
context, a reduction to 70% of the mean tensile strength (2800 MPa) ence may be explained by the influence of the different cross sections as
was suggested and showed a good correlation between calculated and indicated in Section 2.4.5. Compared to the beams tested in the own
experimental values. To prove this assumption, the reduced tensile

Table 7
Shear tests on beams with textile CFRP shear reinforcement.
Researcher # d a/d ρl El ρw Ew fw fcm
[–] [–] [mm] [–] [%] [MPa] [%] [MPa] [MPa] [MPa]

Bongardt [23] 2 161–167 4.2 0.46 211,100 0.06 250,000 3550 150
Kromoser [9] 6 230–234 3.0–4.0 1.71–2.24 161,000 0.23–0.30 – 1300 136
Kueres 2018 1 226 3.5 0.5 187,000 0.19 180,000 1960 86

#: number of tests; d: effective depth; a/d: shear span-depth ratio ρl: longitudinal reinforcement ratio; El: elastic modulus of longitudinal reinforcement; ρw: shear
reinforcement ratio; Ew: elastic modulus of shear reinforcement; fw: governing tensile strength of the shear reinforcement determined in test; fcm: mean value of the
concrete compressive strength.

8
S. Kueres, et al. Engineering Structures 206 (2020) 110088

2.0 2.0
(a) [24] (E) [10]
¿Eer reinforceG [9] [9]
1.5 concrete Kueres 1.5 Kueres

Vtest / Vcalc [-]

Vtest / Vcalc [-]

1.0 1.0

faiOure E\ concrete
0.5 GeOamination in weE 0.5
with At with At
without At without At
0.0. 0.0.
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Shear reinforcement ratio ȡw* [%] Shear reinforcement ratio ȡw* [%]
Fig. 9. Comparison of calculated and experimental shear strength of beams with textile shear reinforcement (a) and pre-tensioned with CFRP tendons (b).

study (b/bw = 1.6), the beams in [9] and [10] have a significantly more – The governing shear crack of test specimens failed by rupture of the
pronounced top flange (b/bw = 4.3 and 7.5). Due to the bending shear reinforcement was identified as the closest shear crack to the
stiffness of the flange, a spreading of the load distribution of the con- support with a crack angle of about 30°. The measurements of the
centrated loads to a larger area of the web is enabled. Moreover, re- changes of thickness indicate an initiation of the shear crack for-
garding the I-beams used by [9], an additional contribution of the mation between 70 and 80% of the ultimate failure load. The largest
prestressed flanges to the shear resistance may be assumed. This me- increase of thickness change was measured between 95% of the
chanism leads to a reduction of the crack openings and slidings of a failure load and failure.
beam, even for low shear reinforcement ratios. These load increasing – The beneficial effects of prestressing and the additional longitudinal
effects of flanges on shear strength of prestressed beams have already reinforcement of textile shear reinforcement on shear strength can
been identified for steel reinforced prestressed I-beams and are de- be taken into account in a consistent manner in the variable strut
scribed in detail in [17]. Both effects are not considered in the devel- inclination model for members with FRP shear reinforcement pro-
oped shear resistance model since it is exclusively based on the capacity posed in [11]. In this context, the longitudinal strain at mid-depth of
of the compression field in the web of a beam. To extend the model to the web and the inner lever arm are determined based on the dif-
these influences, further investigations regarding the shear strength of ferent types of applied reinforcement and the resulting internal
prestressed I-beams would be required. In [25–27], or [28] a similar forces.
load increasing effect of the flange of T-beams on the shear strength has – The evaluation of the extended variable strut inclination model by
also been reported for non-prestressed beams. However, in Fig. 9(a), no means of own test results and tests reported in literature reveals
significant difference regarding the capacity of the own beam and the good agreement between predictions and experimental results for
beams of [9] can be observed. Due to the limited database of tests, this the non-prestressed beams with textile shear reinforcement and the
effect cannot be finally clarified at this point. own shear tests on prestressed beams. In contrast, the shear re-
sistance of the prestressed beams tested by [9] and [10] is under-
4. Summary and conclusions estimated by up to 25% according to the proposed model. This
difference may be explained by the influence of the different cross
Based on the investigations on the shear behavior of prestressed sections but due to the limited database of tests, the effect could not
concrete members with FRP shear reinforcement presented in this be finally clarified.
paper, the following conclusions can be drawn:
Declaration of Competing Interest
– The present test series allows for the investigation of the influence of
the shear slenderness and of the level of prestressing on the shear
The authors declare that they have no known competing financial
strength of prestressed beams with FRP shear reinforcement. The
interests or personal relationships that could have appeared to influ-
shear resistance of the tested beams decreased with increasing shear
ence the work reported in this paper.
slenderness, whereas with increasing pre-tensioning stress the shear
strength increased. These conclusions correspond to the observa-
tions made by other researchers as reported in [9,10]. Acknowledgements
– For all test specimens, which failed by rupture of the textile shear
reinforcement, the failure occurred at the bent portion of the re- The authors gratefully acknowledge the support received by the
inforcement cage. The test specimens that did not fail by rupture of Federal Ministry of Education and Research (BMBF) of Germany and
the shear reinforcement failed by premature bond slip failure. VDI Technologiezentrum GmbH (project CFC-SYS; support code:
Failure by crushing of the compression struts was not observed in 13N13536), as well as by the project partners CG TEC GmbH, Max Bögl
the test series. Bauservice GmbH & Co. KG and solidian GmbH.

Appendix A

This appendix presents the derivation of Eq. (3), which was applied for the theoretical investigations presented in this paper to calculate the
inclination of the compression strut. Further information are also provided in [11].
In 1995, SIGRIST [19] derived the following equation to determine the inclination of the compression struts of a compression field (Fig. 7) based on
investigations on steel reinforced concrete beams:

⎛ 1 ⎞ ε ·E
tan3 θ ·⎜1 + − cotθ − x c = 0
ρy ·α e ⎟ τxy (A1)
⎝ ⎠

9
S. Kueres, et al. Engineering Structures 206 (2020) 110088

where ρy is the geometric shear reinforcement ratio aw/bw, αe denotes the ratio of the elastic moduli Ew/Ec and τxy is the shear stress. In [2], Eq. (A1) is
further simplified to allow for an analytical solution for the use within the shear design of concrete beams with FRP reinforcement. A transformation
of Eq. (A1) yields the following expression:

1 ⎛ 4 ⎛ 1 ⎞ ⎞ ε ·E
· tan θ ·⎜1 + −1 = x c
tanθ ⎜ ρy ·α e ⎟ ⎟ τxy (A2)
⎝ ⎝ ⎠ ⎠
where the constants 1 and −1 are negligible due to minor importance compared to the term 1/(ρy·αe). If now ρy = aw·bw and τxy = V/(z·bw) are
considered, the simple equations given in Eqs. (A3) or (A4) are determined to calculate the angle θ.
εx · z · a w · E w
tan3 θ =
V (A3)
1/3
V
cotθ = ⎛ ⎜
⎞ ⎟

⎝ εx · z · a w · E w ⎠ (A4)

References [15] Rempel S, Kulas C, Will N, Bielak J. Extremely light and slender precast pedestrian-
bridge made out of textile-reinforced concrete (TRC). High Tech Concrete: Where
Technology and Engineering Meet, Springer Verlag. Cham: Switzerland; 2018. p.
[1] Fédération internationale du béton. FRP reinforcement in RC structures. fib bulletin 2530–7.
No. 40; 2007. [16] EN 1992-1-1:2004: Eurocode 2: Design of concrete structures – Part 1-1: General
[2] Kurth MC. Zum Querkrafttragverhalten von Betonbauteilen mit rules and rules for buildings. Incl. Corrigendum 1: EN 1992-1-1:2004/AC:2008,
Faserverbundkunststoff-Bewehrung. Doctoral dissertation. RWTH Aachen incl. Corrigendum 2: EN 1992-1-1:2004/AC:2010, incl. Amendment 1: EN 1992-1-
University; 2012. [in German]. 1:2004/A1:2014.
[3] Oller E, Marí A, Bairán JM, Cladera A. Shear design of reinforced concrete beams [17] Rupf MM. Querkraftwiderstand von Stahlbeton- und Spannbetonträgern mittels
with FRP longitudinal and transverse reinforcement. Compos B Eng Spannungsfeldern. Doctoral dissertation, EPFL Lausanne; 2014 [in German].
2015;74:104–22. [18] Cladera A, Marí A, Bairán JM, Ribas C, Oller E, Duarte N. The compression chord
[4] Razaqpur AG, Spadea S. Shear strength of FRP reinforced concrete members with capacity model for the shear design and assessment of reinforced and prestressed
stirrups. J Compos Construct 2014;19(1). 04014025-1–15. concrete beams. Structural concrete 2016;17(6):1017–32.
[5] Shehata EF. Fibre-reinforced polymer (FRP) for shear reinforcement in concrete [19] Sigrist V. Zum Verformungsvermögen von Stahlbetonträgern (Vol. 210). Doctoral
structures. Doctoral dissertation. University of Manitoba; 1999. dissertation, ETH Zurich; 1995 [in German].
[6] ACI 440.1R-06. Guide for the design and construction of structural concrete re- [20] Sigrist V. Generalized stress field approach for analysis of beams in shear. ACI Struct
inforced with FRP bars. American Concrete Institute (ACI); 2006. J 2011;108(4):479–87.
[7] JSCE Concrete Engineering Series No. 23. Recommendation for design and con- [21] prEN 1992-1-1:2018. Eurocode 2: Design of concrete structures – Part 1-1: General
struction of concrete structures using continuous fiber reinforcing materials. Japan rules and rules for buildings, bridges and civil engineering structures. Final draft of
Society of Civil Engineers (JSCE); 1997. April 2018 by the Project Team SC2.T1; 2018.
[8] CSA S806-12. Design and construction of building structures with fibre-reinforced [22] Ruiz MF, Muttoni A. On development of suitable stress fields for structural concrete.
polymers. Canadian Standards Association (CSA); 2012. ACI Struct J 2007;104(4):495.
[9] Kromoser B, Huber P, Preinstorfer P. Experimental study of the shear behavior of [23] Kaufmann W. Strength and deformations of structural concrete subjected to in-
thin walled CFRP reinforced UHPC structures. In: Conf. Proc. of the 5th plane shear and normal forces. Report No. 234, Institute of Structural Engineering,
International fib Congress, Australia: Melbourne; 2018. ETH Zurich; 1998.
[10] Nabipaylashgari M. Shear strength of concrete beams pre-stressed with CFRP [24] Bongardt C. Experimentelle und theoretische Untersuchungen zum
cables. Master thesis. University of Manitoba; 2012. Querkrafttragverhalten getränkter textiler Bewehrungselemente für Betonbauteile.
[11] Kueres S, Hegger J. Variable strut inclination model for shear design of FRP re- Master thesis, RWTH Aachen University; 2016 [in German].
inforced concrete members with shear reinforcement. Eng Struct 2019. https://doi. [25] Herbrand M. On the development of shear strength models for reinforced and
org/10.1016/j.engstruct.2019.110154. prestressed concrete beams. Doctoral dissertation, RWTH Aachen University; 2017.
[12] Rempel S. Zur Zuverlässigkeit der Bemessung von biegebeanspruchten [26] Giaccio C, Al-Mahaidi R, Taplin G. Experimental study on the effect of flange
Betonbauteilen mit textiler Bewehrung. Doctoral dissertation. RWTH Aachen geometry on the shear strength of reinforced concrete T-beams subjected to con-
University; 2018. [in German]. centrated loads. Can J Civ Eng 2002;29(6):911–8.
[13] Perse S, Hegger J. Experimental investigations on bond behavior of pre-tensioned [27] Placas A, Regan PE. Shear failure of reinforced concrete beams. ACI Struct J
carbon fiber reinforced polymer tendons. In: Conf. Proc. of 12th International PhD 1971;68(10):763–73.
Symposium in Civil Engineering, Prague, Czech Republic, 2018. [28] Cladera A, Marí A, Ribas C, Bairán J, Oller E. Predicting the shear–flexural strength
[14] Kueres S. Analysis and design of concrete beams with pre-tensioned CFRP re- of slender reinforced concrete T and I shaped beams. Eng Struct 2015;101:386–98.
inforcement. Doctoral dissertation. RWTH Aachen University; 2019.

10