You are on page 1of 11

Received: 17 March 2018 Revised: 15 May 2018 Accepted: 19 June 2018

DOI: 10.1002/suco.201800079

TECHNICAL PAPER

Influence of steel, glass and polymer fibers on the cracking


behavior of reinforced concrete beams under flexure
Antonio Conforti1 | Raúl Zerbino2 | Giovanni A. Plizzari1

1
Department of Civil, Environmental,
Architectural Engineering and Mathematics The use of fibers can improve the behavior at serviceability limit states (crack and
(DICATAM), University of Brescia, Italy deflection control) and ultimate limit states (bearing capacity) of reinforced con-
2
Department of Civil Engineering, La Plata crete (RC) elements under flexure. Fibers reinforcement provides a postcracking
National University, La Plata, Argentina
resistance, leading RC element to have a more diffused crack patterns characterized
Correspondence
by narrower and more closely spaced cracks. Some doubts are instead related to
Antonio Conforti, Department of Civil,
Environmental, Architectural Engineering and RC element ductility, which can be affected by crack localization after rebar yield-
Mathematics (DICATAM), University of Brescia, ing. However, most of experiments present in literature relate only to elements in
Italy. steel fiber reinforced concrete with significant residual strengths (fR,1 and fR,3
Email: antonio.conforti@unibs.it
greater than 3.5–4.0 MPa). This paper aims to evaluate the influence of different
fiber type (steel, glass, or polymer macrofibers) on the cracking and strength capac-
ity of RC beams under flexure by using a broad range of Fiber Reinforced Concrete
(FRC) toughness (1.6 ≤ fR,1 ≤ 5.1 MPa and 0.8 ≤ fR,3 ≤ 4.5 MPa). Twenty-one
small scale RC beams with a typical value of longitudinal reinforcement ratio
(0.87%) were tested under flexure. Crack and deflection control, as well as bearing
capacity and crack localization were evaluated as a function of FRC toughness.
Finally, results were compared against mean crack spacing and strength capacity
predictions of fib Model Code 2010.

KEYWORDS

crack control, fiber reinforced concrete, flexure, glass macrofibers, polymer


macrofibers, steel fibers

1 | INTRODUCTION • at SLS, fibers can improve the crack control1–6 and


deflection control7,8 by transferring stresses at crack (ten-
According to building codes of different worldwide Coun- sion softening), leading to a more diffused crack pattern
tries, fiber reinforced concrete (FRC) is now a common characterized by narrower and more closely spaced
material for structural elements where fibers can be added to cracks as compared to RC beams without fibers2,3;
conventional reinforcement or substitute part of the total • at ULS, fibers can be mainly used as shear
amount of rebars due to the enhanced postcracking tensile reinforcement,9–12 as well as to increase in some cases
resistance of concrete. the flexural bearing capacity.7,13,14 In fact, fiber effect on
In reinforced concrete (RC) beams, fibers can be used as
the flexural bearing capacity are either minor7 or impor-
distributed reinforcement to enhance the response at both
tant14 as a function of FRC toughness-to-longitudinal
serviceability limit states (SLS) and ultimate limit states
reinforcement ratio adopted. Some doubts are instead
(ULS). In particular:
related to the possible negative influence of fibers on
Discussion on this paper must be submitted within two months of the print flexural beam ductility, due to the crack localization (one
publication. The discussion will then be published in print, along with the flexural crack widen more than the other cracks do).15,16
authors' closure, if any, approximately nine months after the print publication.

Structural Concrete. 2019;20:133–143. wileyonlinelibrary.com/journal/suco © 2018 fib. International Federation for Structural Concrete 133
134 CONFORTI ET AL.

Another important aspect that can be improved in RC


beams by fibers is the bond between rebar and concrete due
to the improved splitting crack control,17 leading to
enhanced anchorages and laps. However, to get these bene-
fits, slightly different reinforcement detailing should be
applied to RC beams with fibers (especially in terms of spac-
ing of bars and fiber-to-aggregate ratio) in order to guarantee
a satisfactory placed and compacted of concrete.
The fib Model Code 201018 (MC2010) introduced FRC
as structural material, providing design rules to properly take
into account fiber effects. Since FRC is treated as a compos-
ite material in which different fibers can be used (steel, poly-
mer, glass, or others), the proposed design method is based
on postcracking residual strengths. In particular, fR,1 and fR,3
performance parameters, as obtained from three-point bend-
ing tests according to EN 14651,19 are considered. In paral-
lel to MC2010, other guidelines and codes are under
development in different countries (ACI report 544.1R-96,20
CNR DT 204,21 DAfStb22), as well as FRC is going to be FIGURE 1 Geometry and reinforcement details of specimens
included in the new Eurocode 2.23
Unless the shear behavior, in which several articles on to CIRSOC 20125) were produced for determining the mate-
polymer macrofibers9,12 can be found in literature, the exper- rial mechanical properties.
imental/analytical studies mentioned above (from which
code formulations were developed) are mainly related to 2.1 | Specimen geometry and reinforcement details
steel fibers. Consequently, results on different fiber types are Small scale beams of 150 × 150 × 900 mm having a longi-
still missing, as well as only the range of postcracking per- tudinal reinforcement ratio (ρ) of 0.87% (210 rebars with a
formances generally given by steel fibers was studied so far clear concrete cover [c] of 25 mm) were tested (Figure 1).
(fR,1 and fR,3 ranging between 3.5 and 7 MPa). This under- The adopted reinforcement ratio is significant for practice
lines the need of evaluating the influence of other fiber types (values of ρ ranging between 0.8 and 1.0% are generally
and FRC with lower postcracking performances on the flex- used). Stirrups were introduced near the supports to prevent
ural response of RC beams. In fact, different fibers than steel a shear failure; the first stirrup was placed at half times the
one (i.e., glass or polymer ones) are generally able to impart effective depth (d) from point loads to avoid influence on
similar or lower toughness to concrete, as soon as long-term sample crack pattern in the middle-third of the beam. It
behavior of fibers is guaranteed.24 should be also underlined that the adopted sample geometry
Within this framework, the research work presented is expected to have a similar fiber orientation of EN 1465119
herein investigate the influence of three different fiber beams, since cross section and mold geometry are identical;
types (steel, glass and polymer fibers that lead to have this allows to better relate the sample response to the mate-
1.6 ≤ fR,1 ≤ 5.1 MPa and 0.8 ≤ fR,3 ≤ 4.5 MPa) on the flex- rial toughness.
ural response of RC beams with a typical value of longitudi-
nal reinforcement ratio (0.87%). Particular attention was 2.2 | Materials and mixture proportions
devoted to the different stages of crack development. Experi-
mental results allowed to quantify the effects given by each One plain concrete (CONTROL series) and six FRC (S-25,
fiber at both SLS and ULS, even if it was observed that this S-50, G-6, G-12, P-5, and P-10 series) were produced. All of
influence is only related to the provided FRC toughness and them were prepared from the same base concrete mixture
not to fiber type. Finally, the experimentally determined proportion. Three types of fibers were used: hooked-end
mean crack spacing and bearing capacity were compared steel fibers (S), glass macrofibers (G) and polymer (syn-
against the predictions of MC2010. thetic) macrofibers (P). The main fiber characteristics are
given in Table 1.
The base concrete mixture proportions (by weight) were
2 | EX PER IM ENT AL PROGRA M 1:2.25:2.16:0.41 of Portland cement: natural siliceous sand:
12 mm maximum size granitic crushed stone: water. The
Twenty-one RC beams (three samples for any concrete type) amount of a high-range polycarboxilate-based superplasticizer
were cast and tested under flexure. In addition, standard was varied in each concrete in order to obtain a slump of
beams (according to EN 1465119) and cylinders (according 60  10 mm. Each fiber type was added to concrete in two
CONFORTI ET AL. 135

TABLE 1 Characteristics of steel (S), glass (G) and polymer the age of testing. The nominal stress at limit of proportion-
(P) macrofibers
ality (fL) and the residual strengths at crack mouth opening
Fiber designation S G P displacement (CMOD) of 0.5 and 2.5 mm (fR,1 and fR,3) were
Type Steel Glass Polypropylene obtained. The ratios fR,1/fL and fR,3/fR,1, are also shown in
Shape Hooked-end Crimped Embossed Table 3 to provide indication of postcracking trend between
Length, l [mm] 50 36 58 0.5 and 2.5 mm. The characteristic values of these ratios are
Diameter, ϕ [mm] 1.00 0.54 0.67 adopted by MC2010 to verify the minimum requirements for
Aspect ratio, l/ϕ 50 67 86
structural use of FRC (fR,1k/fLk > 0.4 and fR,3k/fR,1k > 0.5).
Tensile strength [MPa] >1,100 >1,700 >640
Considering fR,jk = fR,j  (1–1.64CV) (fib MC201018),
Elastic modulus [GPa] 210 72 6.8
unless series G-6 and P-5, the different FRC adopted could
Density [kg/m3] 7,850 2,680 910
be considered at ULS verifications for MC2010.
It can be observed that the concrete compressive strength
different dosages; series designations and fiber contents are is almost similar between different series (mean value
shown in Table 2. FRC were identified in accordance to the 43.6 MPa with overall CV of 0.05), allowing in the follow-
type and content of fibers (S-25 and S-50, G-6 and G-12, P-5, ing a direct comparison between series (without normaliza-
and P-10), while CONTROL represents samples without tion to concrete compressive strength).
fibers. CONTROL series is adopted as reference for evaluating Regarding the postcracking behavior of the different
the fiber influence on the flexural behavior of RC beams. FRC, Figure 2 shows the mean experimental curves from
For any concrete type, three small scale RC beams of EN 14651 tests, in terms of nominal stress versus CMOD.
150 × 150 × 900 mm (Figure 1), six beams of When considering the studied FRC, some important evi-
150 × 150 × 550 mm (according to EN 1465119) and six dences should be highlighted. The limit of proportionality is
cylinders of 100 × 200 mm (according to CIRSOC 20125) not significantly affected by the presence of fibers, as con-
were cast. All specimens were cured in moist room for firmed also by the fL/fc ratio. After cracking, very different
28 days and then remain in laboratory indoor up to testing. postcracking curves can be observed, depending on the fiber
Testing ages were between 60 and 90 days in order to mini- type and amount. Steel fibers showed a softening and hard-
mize the variation of concrete mechanical properties during ening behavior under flexure in case of 25 and 50 kg/m3,
the testing period (both EN14651 beams and cylinders were respectively. Glass fibers exhibited a continuous softening
tested at the same age of RC beams). behavior, while polymer macrofibers showed a softening
behavior with a sharp load drop after the peak load, followed
2.3 | Mechanical characterization test results by an increment of the residual strength. Therefore, a broad
Table 3 shows the mean values (coefficients of variations range of FRC residual strengths were considered
[CV] in brackets) of compressive and flexural test results at (1.6 ≤ fR,1 ≤ 5.1 MPa and 0.8 ≤ fR,3 ≤ 4.5 MPa). This
underlines the importance of the present research where dif-
TABLE 2 Series designation and fiber contents ferent fibers, as well as low toughness FRC are studied. In
Fiber volume fact, as already underlined, most studies present in literature
Series Fiber type Fiber content [kg/m3] fraction [%] dealing with the cracking behavior in FRC elements are
CONTROL — 0 0 referred only to steel fiber reinforced concrete (SFRC) char-
S-25 S 25 0.32 acterized by greater values of postcracking residual
S-50 S 50 0.64
strengths.
G-6 G 6 0.22
Rebars were characterized according to EN 15630–126
G-12 G 12 0.44
by testing 600 mm long pieces. The overall mean values of
P-5 P 5 0.55
fy and fu resulted equal to 469.4 MPa (CV = 0.01) and
P-10 P 10 1.10
632.3 MPa (CV = 0.02) for 6 stirrups; 562.3 MPa

TABLE 3 Compressive and bending properties of concretes

Series fc [MPa] fL [MPa] fR,1 [MPa] fR,3 [MPa] fL/fc [-] fR,1/fL [-] fR,3/fR,1 [-]
CONTROL 42.1 (0.04) 3.71 (0.10) - - 0.09 - -
S-25 43.7 (0.04) 4.51 (0.12) 3.22 (0.10) 2.90 (0.02) 0.11 0.71 0.90
S-50 42.9 (0.04) 4.69 (0.04) 5.08 (0.02) 4.53 (0.11) 0.11 1.08 0.89
G-6 45.5 (0.03) 4.51 (0.02) 1.63 (0.19) 0.76 (0.13) 0.10 0.36 0.47
G-12 44.3 (0.02) 4.96 (0.01) 2.89 (0.09) 1.70 (0.13) 0.11 0.58 0.59
P-5 40.0 (0.07) 3.65 (0.04) 1.62 (0.10) 1.64 (0.17) 0.09 0.44 1.01
P-10 46.8 (0.05) 4.30 (0.07) 2.84 (0.12) 4.02 (0.17) 0.09 0.66 1.42
136 CONFORTI ET AL.

bottom face level of the beam with the aim of evaluating the
extensibility and the sum of cracks openings (wH) in the
middle third of the beam during the test. Particular attention
was also devoted to the evaluation of the crack pattern devel-
opment and crack widths on the flexural span. In this
respect, crack width and crack patterns were captured at dif-
ferent load stages: 50% of the flexural bearing capacity
(p = 40 kN, which corresponds to a steel stress equal to
340 MPa); 70% of the flexural bearing capacity (p = 56 kN,
which corresponds to a steel stress equal to 480 MPa) and
rebar yielding. Additional measurements were also carried
out after rebar yielding at different levels of mid-span
deflection.
Specimens were monotonically loaded with a rate of
0.3 mm/min; at each load stage, the displacement was held
while the cracks were marked and evaluated in terms of
FIGURE 2 Nominal stress versus CMOD mean curves for the different
concretes crack width. A comparator and a magnifying glass were
used for that purpose. Each test took about 2 hours.

3 | R E S U L T S AN D D IS C US S IO N

All RC beams with and without fibers showed a flexural fail-


ure (the compression zone reached its maximum strain after
rebar yielding). Table 4 summarizes the mean values (CV in
brackets) of the main experimental results obtained. The load
at flexural cracking (Pcr), the net mid-span deflection at
p = 40 kN (δP40), the maximum flexural crack width (wmax)
measured at 40 kN and 56 kN, the mean crack spacing at
crack stabilized stage (sr), the maximum load (Pmax) and the
maximum net mid-span deflection (δu) are reported. In addi-
tion, since in some cases crack localization occurred after
rebar yielding, the individual values of the net mid-span
deflection at the crack localization initiation (δCL) are shown.
These parameters will be used in the next sections to analyze
the fiber influence on the flexural behavior. It should be also
FIGURE 3 Four point loading test: instrumentation details and test set-up noticed that CV are in general small and comparable
between series, evidencing that the experimental results are
consistent.
(CV = 0.03) and 690.7 MPa (CV = 0.01) for 10 Figure 4 shows the experimental load - net mid-span
reinforcing bars. deflection mean curves of series incorporating steel (S-25
and S-50), glass (G-6 and G-12) and polymer (P-5 and P-10)
2.4 | Test set-up and instrumentation of RC beams fibers compared against the mean response of beams pre-
pared without fibers (CONTROL). The low dispersion of the
Figure 3 shows a general view and the instrumentation experimental results (Table 4) allowed to consider the mean
details of RC beam tests. Four point loading configuration curves. It should be also noticed that in Figure 4 the results
was applied. The adopted net span was 840 mm, while the are only plotted up to a net mid-span deflection of 10 mm to
distance between the two point loads (flexural span) was make more visible the sample responses at cracking stages
equal to 280 mm (Figure 1). In order to ensure a suitable before yielding. From Figure 4 the different stages of crack-
control, all tests were carried out in displacement control ing development can be defined:
using an INSTRON machine with a loading capacity of
1,000 kN. I. the initial uncracked stage (up to the cracking load Pcr).
The measurement of the net mid-span deflection (δ) was No influence of fiber type and dosage was observed in
made through a LVDT placed on a frame fixed at the neutral this stage, as well as Pcr resulted similar in all concretes
axis over each support. Another LVDT was placed at the (Table 4);
CONFORTI ET AL. 137

TABLE 4 Experimental results of flexural tests on RC beams: mean values and coefficients of variation

Pcr δP40 wmax,P40 wmax,P56 sr δCL Pmax δu


Series [kN] [mm] [mm] [mm] [mm] [mm] [kN] [mm]
CONTROL 13.0 (0.10) 1.43 (0.10) 0.15 (0.00) 0.22 (0.13) 88 (0.20) - 77.5 (0.03) 15.7 (0.06)
-
-
S-25 12.9 (0.21) 1.26 (0.04) 0.10 (0.00) 0.17 (0.17) 69 (0.14) 4.8 85.4 (0.04) 15.8 (0.18)
-
6.3
S-50 13.9 (0.02) 1.03 (0.07) 0.07 (0.43) 0.13 (0.22) 62 (0.07) 7.1 87.9 (0.01) 14.1 (0.19)
5.9
4.3
G-6 11.8 (0.06) 1.31 (0.13) 0.10 (0.50) 0.17 (0.15) 71 (0.09) - 79.5 (0.04) 15.5 (0.06)
-
-
G-12 12.7 (0.14) 1.27 (0.05) 0.10 (0.00) 0.15 (0.00) 65 (0.11) 6.9 78.0 (0.02) 13.8 (0.14)
8.8
6.7
P-5 13.3 (0.04) 1.30 (0.07) 0.13 (0.22) 0.22 (0.13) 69 (0.16) - 79.0 (0.05) 15.9 (0.20)
8.0
-
P-10 13.8 (0.11) 1.20 (0.05) 0.13 (0.22) 0.20 (0.00) 63 (0.01) 6.6 82.4 (0.03) 17.6 (0.06)
5.9
7.6

FIGURE 4 Load versus net mid-span deflection curves (up to 10 mm) for RC samples incorporating steel, glass and polymer fibers
138 CONFORTI ET AL.

II. the cracked stage that includes crack formation and sta- polymer macrofibers (Figure 2), especially when they are
bilized crack stage (now further crack occurs). The incorporated in high dosages (small values of residual
presence of fibers changed the crack pattern and the strength for small crack widths (SLS) and greater values of
beam stiffness (slope of the curves) depending on fiber residual strength for larger crack widths (ULS)).
type and dosage by the transmission of residual In the following sections the main experimental results
strength through cracks (for a load of 40 kN the post- (Table 4) will be related to the concrete toughness (Table 3)
cracking stiffness reduction is quantified thorough the with the aim of evaluating a possible relation (regardless
parameter δP40 in Table 4). The stabilized crack stage fiber type and dosage). The wide range of toughness consid-
took place for all samples at a deflection around ered makes possible a deep analysis of the phenomenon.
2.5 mm, underlining as its onset seems not to be influ-
enced by fiber types and FRC toughness (even if a
wide range of material toughness was considered). This 3.1 | Fiber influence at SLS
is in accordance to previous results obtained by Tiberti Figure 5 shows the influence of fibers on the maximum
et al.6 on SFRC tension ties;
crack width (a) and on the mean crack spacing (b) as a func-
III. the steel yielding stage. The load at which the steel
tion of fR,1; the latter is the postcracking parameter of EN
yielding occurred was related to concrete toughness
14651 related to SLS. It is worth mentioning that the crack
(the load increase at increasing concrete toughness).
width associated to fR,1 at the notch tip is around 0.36 mm,27
This is due to the postcracking stresses transfer by
which is greater than the measured crack width on RC
fibers at any concrete crack (tension softening). The net
beams at SLS (between 0.07 and 0.22 mm, Table 4). It
mid-span deflection at which the yielding took place
appears that the maximum crack opening, significant for
was instead similar in all series due to deformation
SLS, is reduced by the presence of fibers. A linear relation
compatibility.
provides a good approximation of the maximum crack width
as a function of fR,1 (the accuracy was quantified by the coef-
It can be also observed a clear influence of steel fibers
both at SLS (in terms of deflection reduction at cracked ficient of determination, R2). This relation seems to be only
stage) and ULS (maximum bearing capacity). This influence slightly influenced by the load level considered (similar line
is higher as the fiber dosage increases (from S-25 to S-50) slopes can be observed in Figure 5). Regarding the fiber type
and it is more significant at ULS as compared to SLS. An it was observed that steel fibers significantly reduce the max-
evident reduction of net mid-span deflection as compared to imum crack opening (S-25 of 25–30%, S-50 of 40–55%);
reference samples (CONTROL) is observed only in case of glass fibers also reduce the maximum crack opening of
S-50 (more details will be given in Section 3.1). Glass fibers about 25–30%, while in case of P fibers the reduction was
exhibited a small influence both in terms of deflection reduc- lower than 10%. This is probably due to the low postcrack-
tion and bearing capacity increment. Finally, although the ing performances of FRC with polymer macrofibers for
polymer fibers show some effects on both postcracking stiff- small crack opening. In fact, even if the values of fR,1 are
ness and maximum bearing capacity, this influence is smal- similar between G and P fibers, for the values of crack width
ler than the one of steel fibers, and unless P-10 at ULS, this observed at SLS in the adopted RC beams (smaller than
influence is similar to the case of glass fibers. This behavior 0.36 mm), G fibers provided higher residual strengths as
is related to the postcracking response that characterizes compared to P ones.

FIGURE 5 Influence of the residual strength fR,1 on maximum crack width (a) and mean crack spacing (b)
CONFORTI ET AL. 139

FIGURE 6 Influence of the residual strength fR,1 on bottom beam strain in the central middle third at different load levels (a) and sample postcracking
deflection (b)

Fibers, as expected, led also to a reduction of mean crack (evaluated as flexural elongation divided by measurement
spacing (Figure 5b) due to the stresses transfer by fibers at length, wH/LH) as a function of fR,1; this deformation reduces
any concrete crack with provoke a reduction of transmission when postcracking strength (fR,1) increases. This reduction is
length (ls,max). This is accordance to previous works on steel significantly smaller to the one on maximum crack opening.
fibers.1–6 This reduction seems to be linearly related to fR,1 In fact, S-50 leads to a reduction of 20% while in case of G
and, therefore, it is not depending on fiber type. This reduc- fibers and P fibers it was around 10%. Therefore, the overall
tion was observed also for small values of fR,1 (e.g., P-5 and strain deformation in flexure is smaller in FRC specimens;
G-6), for example, a fR,1 near 1.5 MPa leads to a reduction this provokes the reduction of postcracking deflection
of about 15%. The highest effects were observed for steel depicted in Figure 6b. Once again, the relation between post-
fibers (22% for S-25 and 30% for S-50), even if a reduction cracking deflection and fR,1 can be approximated as linear,
around 20% was also observed for G and P fibers with the as well as the reduction are comparable to the ones depicted
highest dosage. in Figure 6a. A significant effect of fibers on the postcrack-
Thus, RC beams with fibers are characterized by differ- ing deflection reduction was observed for fR,1 ≥ 3.0 MPa.
ent crack patterns in flexure as compared to RC samples Under this value the reduction is less than 10%, and there-
without fibers, since more cracks but smaller in term of fore it can be neglected. Also Figure 6 shows that this influ-
crack width were observed. Consequently it is important to ence does not depend on fiber type but only on FRC
figure out if the overall flexural deformation along longitudi- performances (values of R2 close to 1 can be also observed).
nal rebars is comparable or not. In this way, Figure 6a shows
the bottom beam strains in the central middle third 3.2 | Fiber influence at ULS
Figure 7 shows the influence of fibers on the maximum bear-
ing capacity of a beam under flexure (ρ = 0.87%) as a func-
tion of fR,3. The latter is the postcracking parameter of EN
14651 related to ULS. In the range of fR,3 considered, it can
be observed a linear relation between these two parameters
with a quite consistent coefficient of determination. For low
values of fR,3 (G-6, G-12 and P-5), fiber contribution is smal-
ler than 5% and thus, it could be neglected. The highest
influence was observed for steel fiber (S-25 + 10%, S-
50 + 13%), while P fibers in the highest amount led to an
influence slightly smaller than 10%. Consequently, only
steel or high amount of polymer fibers can modify the maxi-
mum bearing capacity of RC beams with a longitudinal rein-
forcement ratio around 0.9%.
Flexural behavior at ULS of a RC beams involves also a
discussion of ductility. The latter can be expressed in terms
FIGURE 7 Influence of the residual strength fR,3 on the maximum bearing of displacement by the ratio between δu and δy, where δu is
capacity of RC beams the maximum deflection (at collapse) and δy is the deflection
140 CONFORTI ET AL.

FIGURE 8 Crack development and final crack pattern of a sample of series CONTROL, G-6 and S-50

at rebar yielding; since δy was similar in all series, a direct localization, FRC samples showed greater crack openings as
comparison between the values of δu shown in Table 4 can compared to RC samples. This is completely the opposite as
be carried out. It can be observed that all series were charac- compared to the improved crack control given by fibers at
terized by high and comparable ductility, underlining as for SLS, even if crack control is not required at ULS.
the typical values of longitudinal reinforcement ratio adopted Figure 8 shows also the final crack patterns (after image
in practice, FRC toughness (in the considered range) do not rectification) of a sample of series CONTROL, G-6 and S-
reduce RC beam ductility. Table 4 reports also the deflection 50; it can be clearly observed the high number of cracks in
at crack localization initiate (δCL). It can be observed that the case of FRC samples, as well as the crack localization.
crack localization always took place after rebar yielding and
it is not possible to establish a clear relation between the ini-
tiation of crack localization (δCL) and the concrete tough- 3.3 | Comparison against MC2010 predictions
ness. However, for higher values of concrete toughness, Both mean crack spacing and flexural bearing capacity pre-
crack localization was always observed, while for smaller dictions of MC2010 were evaluated against the experimental
concrete toughness it was observed only in some cases. It results presented herein. The predictions were calculated by
should be noticed that crack localization took place also for assuming strength reduction factors equal to 1 and the mean
FRC characterized by low-concrete toughness. However, values of the material mechanical properties. As mentioned
crack localization did not reduce the ductility in flexure, and above, the fiber orientation in RC beams and EN 14651
the failure was always characterized by concrete crushing beams is supposed to be similar, allowing a direct
well after rebar yielding. As shown in Figure 8, after crack comparison.
CONFORTI ET AL. 141

The mean crack spacing formulation for RC and FRC FRC samples can be observed, being the mean values
elements considers the fiber influence by a decrease of the around 1.0. Consequently, the estimation model adopted in
transmission length (ls,max) as a function of the FRC tough- MC2010 is consistent and reliable, keeping the same degree
ness at SLS (fFTs = 0.45 fR,1). The following expression for of accuracy with increasing FRC toughness.
sr can be derived from MC2010: Finally, in order to evaluate MC2010 formulations
" # against a wider range of experimental results and to improve
1  ðfct − 0:45  fR, 1 Þ them, as well as to deeply study the fiber influence on the
sr, MC2010 = 1:17  ls, max = 1:17  k  c +   ,
4 ρs, ef τb
flexural behavior of RC beams, future works should be car-
ð1Þ ried out considering a wide range of longitudinal reinforce-
where k is an empirical parameter that can be considered ment ratio, clear concrete cover, concrete compressive
equal to 1, ø is the diameter of longitudinal reinforcement, strength, and bar diameter.
ρs,ef is the effective reinforcement ratio, fct is the concrete
tensile strength evaluated according to MC2010, τb is the
mean bond strength between steel and concrete that can be
4 | C O NC LU DING R EMA RKS
assumed as 1.8fct.
Table 5 summarizes the experimental mean crack spacing The influence of steel, glass or polymer macrofibers on the
(see also Table 4), the crack spacing predictions (sr,MC2010) flexural behavior of RC beams was studied by means of
and the ratio between sr and sr,MC2010. It can be noticed that,
small scale samples with a longitudinal reinforcement ratio
a smaller crack spacing is generally predicted by the
of 0.87%. Based on this experimental study, the following
MC2010; in case of RC beams without fibers (CONTROL)
conclusions might be drawn:
this underestimation is around 7%, while for elements incor-
porating fibers this underestimation increase up to a mean
1. The effect of fibers on the flexural behavior of RC
value of 11%. This underestimation seems to increase with
beams is mainly related to FRC toughness, regardless of
concrete toughness increment (higher ratio was observed for
fiber type and amount. This confirm the reliable philoso-
high values of residual properties; see steel fibers or high
dosage of glass fibers). Similar results on SFRC were also phy adopted by MC2010, such that FRC can be consid-
found by Tiberti el al.,6 where the results of a broad experi- ered as a composite material where performance
mental campaign on tensions ties were compared against parameters (i.e., fR,1 and fR,3) govern its mechanical
MC2010 crack spacing formulation. This underlines that this behavior;
formulation may require some improvements. 2. The maximum crack width at SLS can be significantly
The prediction of the flexural bearing capacity according reduced by using steel (30–50%) or glass fibers
to MC2010 (Mu,MC2010) was calculated assuming the yield- (25–30%), while the effect of polymer macrofibers is
ing of rebars (values of fy listed in Section 2.3 were consid- lower (less than 10%) due to their low postcracking per-
ered) and an equivalent rectangular concrete compressive formances for small crack openings;
stress distribution (stress block). In addition, the contribution 3. A mean crack spacing reduction due to fibers was
of fibers was considered in flexure, by a simplified model observed also for FRC with small values of fR,1. The
adopted to compute the ultimate tensile strength fFTu, where: most remarkable influence was found in case of steel
fFTu = fR,3/3. Table 5 summarizes also the experimental flex- fibers (22% for S-25 and 30% for S-50), even if a reduc-
ural strength Mu, Mu,MC2010 and the ratio between Mu and tion of about 20% was also observed for glass and poly-
Mu,MC2010. Good predictions both in case of CONTROL and mer macrofibers in the highest dosage;

TABLE 5 Mean crack spacing and flexural strength predictions of fib Model Code 2010

Specimen designation sr [mm] Mu [kNm] sr,MC2010 [mm] Mu,MC2010 [kNm] sr/sr,MC2010 Mu/Mu,MC2010
CONTROL 88 10.8 82 10.1 1.07 1.07
S-25 69 12.0 58 11.5 1.18 1.04
S-50 62 12.3 44 12.2 1.40 1.01
G-6 71 11.1 70 10.5 1.01 1.06
G-12 65 11.0 61 11.0 1.06 1.00
P-5 69 11.1 69 10.8 1.00 1.02
P-10 63 11.5 62 12.0 1.01 0.96
Mean FRC series - - - - 1.11 1.01
CV FRC series - - - - 0.14 0.04
142 CONFORTI ET AL.

4. The fiber influence on the reduction of the postcracking


LH measurement length of the horizontal
deflection becomes significant for fR,1 ≥ 3.0 MPa,
instrument
which are values generally achievable by steel fibers;
ls,max transmission length
5. Steel and polymer macrofibers (in high amount) can
Mu flexural strength
enhance the flexural bearing capacity of RC beams, Mu,MC2010 predicted flexural strength according to fib
while the effects of glass fibers can be neglected since it Model Code 2010
leads to a strength increment smaller than 5%;  bar diameter
6. FRC toughness (in the considered range) does not P load
reduce RC beam ductility, even if a crack localization Pcr load at flexural cracking
occurred after rebar yielding in the majority of FRC Pmax maximum load
samples. No clear relation between the initiation of crack R2 coefficient of determination
localization and FRC toughness was observed; sr mean crack spacing
7. MC2010 approach for evaluating the flexural strength of sr,MC2010 predicted mean crack spacing according to fib
FRC beams is consistent and reliable, while mean crack Model Code 2010
spacing predictions seem to be characterized by a pro- w crack width
gressive increasing underestimation for FRC toughness wH flexural elongation on central middle third from
increase. the horizontal instrument
wmax maximum flexural crack width
Further tests should carried out by varying the other prin- δ net mid-span deflection
ciple variables in order to obtain more data that allow, where δCL net mid-span deflection at crack localization
needed, to improve the current MC2010 prediction models. initiation
δP40 net mid-span deflection at p = 40 kN
δu maximum net mid-span deflection
ACKNOWLEDGMENTS εs bottom beam strain calculated as wH/LH
The Authors would like to give their appreciation to Engi- ρ longitudinal reinforcement ratio
neers Michele Fasciolo and Maria Celeste Torrijos, as well ρs,ef effective reinforcement ratio (=As/Ac,ef)
as to the technicians Anabela Gerez and Pablo Bossio of τb mean bond stress between steel and concrete
LEMIT-CIC, for the assistance in performing the experimen- ϕ fiber diameter
tal program. A special acknowledgement goes also to Prof.
Giuseppe Tiberti for its precious suggestions. OR CID

Antonio Conforti https://orcid.org/0000-0003-2796-7409

5 | NOTATIONS
REFERENCES
Ac,ef effective area of concrete in tension 1. Abrishami HH, Mitchell D. Influence of steel fibers on tension stiffening.
As area of reinforcement ACI Struct J. 1997;94:769–773.
c clear concrete cover 2. Noghabai K. Effect of tension softening on the performance of concrete struc-
tures. Experimental, analytical and computational studies [PhD thesis]. Divi-
CMOD crack mouth opening displacement sion of Structural Engineering, Lulea University of Technology; 1998, p. 186.
d beam effective depth 3. Bischoff PH. Tension stiffening and cracking of steel fiber-reinforced con-
fc mean value of the cylinder compressive con- crete. J Mater Civ Eng. 2003;15(2):174–182.
4. Fields K, Bischoff PH. Tension stiffening and cracking of high-strength rein-
crete strength
forced concrete tension members. ACI Struct J. 2004;101:447–456.
fct mean value of the axial tensile concrete 5. Chiaia B, Fantilli AP, Vallini P. Evaluation of crack width in FRC structures
strength and application to tunnel linings. Mater Struct. 2009;42:339–351. https://
fFtu ultimate residual strength doi.org/10.1617/s11527-008-9385-7.
6. Tiberti G, Minelli F, Plizzari GA. Cracking behavior in reinforced concrete
fL mean value of limit of proportionality members with steel fibers: A comprehensive experimental study. Cem Concr
fR,1 mean value of residual flexural tensile strength Res. 2015;68:24–34.
corresponding to CMOD = 0.5 mm 7. Meda A, Minelli F, Plizzari GA. Flexural behaviour of RC beams in fibre
fR,3 mean value of residual flexural tensile strength reinforced concrete. Compos Part B Eng. 2012;43(8):2390–2937.
8. Amin A, Foster SJ, Kaufmann W. Instantaneous deflection calculation for steel
corresponding to CMOD = 2.5 mm fibre reinforced concrete one way members. Eng Struct. 2017;131:438–445.
fR,j residual flexural tensile strength corresponding 9. Altoubat S, Yazdanbakhsh A, Rieder KA. Shear behavior of
to CMOD = CMODj macro-synthetic fiber-reinforced concrete beams without stirrups. ACI Mater
J. 2009;106:38–389.
fR,jk characteristic value of fR,j
10. Dinh HH, Parra-Montesinos GJ, Wight J. Shear behaviour of steel
h beam height fibre-reinforced concrete beams without stirrup reinforcement. ACI Struct J.
l fiber length 2010;107:597–606.
CONFORTI ET AL. 143

11. Cuenca E, Echegaray-Oviedo J, Serna P. Influence of concrete matrix and AUTHOR'S BIOGRAPHIES
type of fiber on the shear behavior of self-compacting fiber reinforced con-
crete beams. Compos Part B Eng. 2015;75:135–147.
12. Ortiz-Navas F, Navarro-Gregori J, Leiva-Herdocia GE, Serna P, Cuenca E. Antonio Conforti
An experimental study on the shear behaviour of reinforced concrete beams Ph.D., Postdoctoral Fellow
with macro-synthetic fibres. Constr Build Mater. 2018;169:888–899. https:// Department of Civil, Environmental,
doi.org/10.1016/j.conbuildmat.2018.02.023.
13. Caratelli A, Meda A, Rinaldi Z, Romualdi P. Structural behaviour of precast Architectural Engineering and Math-
tunnel segments in fiber reinforced concrete. Tunn Undergr Space Technol. ematics (DICATAM)
2011;26:284–291. https://doi.org/10.1016/j.tust.2010.10.00. University of Brescia,
14. Conforti A, Tiberti G, Plizzari GA, Caratelli A, Meda A. Precast tunnel seg-
ments reinforced by macro-synthetic fibers. Tunn Undergr Space Technol.
Italy
2017;63:1–11. https://doi.org/10.1016/j.tust.2016.12.005. antonio.conforti@unibs.it
15. Schumacher P. Rotation capacity of self-compacting steel fibre reinforced
concrete [PhD thesis]. TU-Delft; 2006. Raúl Zerbino
16. Werzbrger S, Karinski YS, Dancygier AN. Quantification of cracking locali- Associate Professor and CONICET
zation in fibre-reinforced concrete beams. IOP Conf Series: Materials Sci- Researcher at LEMIT-CIC
ence and Engineering. 2017;246:012024.
17. Bigaj Van-Vliet A. Bond of deformed reinforcing steel bars embedded in Department of Civil Engineering
steel fiber reinforced concrete—State of the Art. Technical Report, La Plata National University, La
TU-Delft; 2001, p. 65. Plata, Argentina
18. Fédération Internationale du Béton, fib Model Code 2010—Final draft,
vol. 1; 2012, p. 350; vol. 2, p. 370.
zerbino@ing.unlp.edu.ar
19. EN 14651. Precast concrete products—Test method for metallic fibre
Giovanni A. Plizzari
concrete—Measuring the flexural tensile strength CEN, Brussels, Belgium:
European Standard, 2005. Professor of Structural Engineering
20. ACI Committee 544. Report on Fiber Reinforced Concrete. ACI Report Department of Civil, Environmental,
544.1R-96. American Concrete Institute; 1996, p. 66. Architectural Engineering and Math-
21. CNR DT 204. Guidelines for the design, construction and production control
of fibre reinforced concrete structures. Roma, Italy: National Research ematics (DICATAM)
Council of Italy, 2006;p. 59. University of Brescia,
22. Deutscher Ausschuss für Stahlbeton (DAfStb) Guideline. German Commit- Italy
tee for Reinforced Concrete, Steel fibre reinforced concrete; design and con-
struction, specification, performance, production and conformity, execution
giovanni.plizzari@unibs.it
of structures. Berlin, Germany: Beuth Verlag GmbH, 2012;p. 48.
23. CEN TC 250/SC2/WG1/TG2. SFRC-steel fibre reinforced concrete. Annex
to EN 1992-1; 2016 (under development).
24. Serna P, Llano-Torre A, Cavalaro SHP. Creep behaviour in cracked sections
of fibre reinforced concrete. Proceedings of the International RILEM Work-
shop FRC-CREEP, RILEM Bookseries, Springer, The Netherlands, 2017;
14, ISSN: 2211-0844.
25. CIRSOC 201. Reglamento Argentino de Estructuras de Hormigón. Argen-
How to cite this article: Conforti A, Zerbino R,
tina: INTI – CIRSOC, 2005;p. 482. Plizzari GA. Influence of steel, glass and polymer
26. EN 15630-1. Steel for the reinforcement and prestressing of concrete—Part fibers on the cracking behavior of reinforced concrete
1: Test methods CEN, Brussels, Belgium: European Standard; 2004.
beams under flexure. Structural Concrete. 2019;20:
27. Conforti A, Minelli F, Plizzari G, Tiberti G. Comparing test methods for the
mechanical characterization of fiber reinforced concrete. Struct Concr. 2018; 133–143. https://doi.org/10.1002/suco.201800079
19:656–669. https://doi.org/10.1002/suco.201700057.

You might also like