Review article

Bond and size effects on the shear capacity of RC beams without stirrups
Jacinto R. Carmona

, Gonzalo Ruiz
E.T.S. de Ingenieros de Caminos, Canales y Puertos, Universidad de Castilla-La Mancha, 13071 Ciudad Real, Spain
a r t i c l e i n f o
Article history:
Received 26 July 2013
Revised 27 December 2013
Accepted 31 January 2014
Keywords:
Shear strength
Size effect
Bond effect
a b s t r a c t
The paper presents a model which allows studying the inuence of bond between the reinforcing bars
and the concrete matrix and the size effect on the evaluation of shear strength in reinforced concrete
beams without stirrups. The formulation assumes that shear failure is caused by the propagation of ex-
ural cracks. When the crack length reaches a certain depth, the so-called critical depth, the section col-
lapses. This depth depends on the position of the section being studied, the external load, the beam
boundary conditions and geometry. Non-linear concepts of Fracture Mechanics are used to model con-
crete behavior in tension during the crack propagation. Size effect is reproduced through Bazants law
(Bazant and Pfeiffer, 1987). Bond slip is considered by a rigid-plastic bondslip curve. The results of
the model can provide an understanding of the inuence of steel-to-concrete bond on the shear strength
and the size effect exhibited in test results and its asymptotic behavior. All these topics are of the utmost
importance to concrete technology, yet they are not satisfactorily dealt with by construction codes.
2014 Elsevier Ltd. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2. Modeling shear crack propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.1. Plain concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.2. Reinforced concrete. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3. Failure criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4. Crack shape. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.1. Model response and experimental validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.2. Size effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.3. Bond effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
1. Introduction
The evaluation of the load that causes shear failure due to diag-
onal tension in RC elements is a problem that has yet to be satisfac-
torily resolved. A consensus over a mechanical model, which may
simply and reasonably explain the behavior of the RC elements
experiencing this type of failure and the inuence of the bond be-
tween concrete and steel, has not been reached as of the present
time. Nevertheless, the interest in this subject is apparent from
the hundreds of publications written about it over the last fty
years. Moreover, the ability of the proposed models to predict
has gradually increased. The models to determine the failure load
due to diagonal tension in elements without stirrups have evolved
from hypotheses based on empirical statistics [13] to truss mod-
els based on plasticity [48].
A new perspective for analyzing the problem was introduced by
Reinhardt in the 1980s [9]. He stated that models and formulas for
determining diagonal traction failure should be based on Fracture
Mechanics (FM). Indeed, the brittle nature of diagonal tension fail-
http://dx.doi.org/10.1016/j.engstruct.2014.01.054
0141-0296/ 2014 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +34 616021731.

E-mail address: Jacinto.rc@gmail.com (J.R. Carmona).
Engineering Structures 66 (2014) 4556
Contents lists available at ScienceDirect
Engineering Structures
j our nal homepage: www. el sevi er. com/ l ocat e/ engst r uct
ure, together with the size effect observed in tests [10,11] and the
fact that failure is associated with crack propagation through the
concrete element [12], suggest that shear failure can be studied
through theories generated within the framework of Fracture
Mechanics. In fact, several models have been proposed to increase
our understanding of the causes and variables that intervene in the
problem. Special emphasis must be placed on the works by
Gustafsson [13] and Hillerborg and Gustafsson [14], who applied
the cohesive crack model, formulated by Hillerborg et al. [15] to
the study of diagonal tension failure and the works about size ef-
fect on shear failure of RC beams performed by Bazant and Yu
[16,17]. Different formulations have also been proposed within
the eld of Linear Elastic Fracture Mechanics for the purpose of
providing simple mechanical models that do not need to use nite
elements. Among them, we highlight the works by Jenq and Shah
[18] and by Carpinteri et al. [19]. Both models introduce the idea
that failure is caused by a exural crack that propagates in the
beam, which demonstrates the existing relationship between the
failure due to bending and diagonal tension (shear). This relation-
ship is disregarded by most standards since both phenomena are
addressed independently.
Another important landmark in the study of shear failure was
the work done by Kani and Wittkopp [20]. Based on experimental
observations, they noted that the type of failure developed in rect-
angular RC beams without web reinforcement was strongly related
to the shear span to depth ratio. They detected a minimum in the
curve shear strength against slenderness, the well known Kanis
shear valley [10]. Related to this effect, Niwa [21,22], in a nite-ele-
ment numerical study, reported the existence of a minimum in the
failure shearing force when the crack initiation position is changed
along the shear span. The same conclusion was drawn by Carpinte-
ri et al. [23]. They used an analytical LEFM model to show failure
mode transitions by varying the controlling non-dimensional
parameters. They demonstrated that shear cracks initiating around
the middle of the shear span need less load to grow in an unstable
manner than cracks near the beam midspan (exural cracks) or
than cracks near the support. The presence of stirrups avoids the
unstable crack growth and changes the collapse mode from shear
to bending. Thus, the connection between crack propagation and
element collapse is key to model the shear strength.
Crack opening and propagation depend on bond conditions. For
that reason, an understanding of its effect on crack propagation is
necessary to evaluate the failure load. Within the framework of
Fracture Mechanics, the inuence of bond between concrete and
steel has been studied for lightly-reinforced beams failing by bend-
ing [2427], but there are no references of previous analytical
models implementing the effect of the inuence of steel-to-con-
crete bond on the shear behavior. Thus, there was a need for a
study that would cover such a topic.
Therefore, Fracture Mechanics provides a sound theoretical
framework in order to study the failure by diagonal tension (shear
failure), its size effect and the inuence of bond properties. A cor-
rect analysis of size effect is important, since most of the tests used
to validate shear strength of beams were performed on small spec-
imens reinforced with adherent bars. Therefore, they cannot be di-
rectly used to derive empirical formulas because the shear
behavior is bond and size dependent. Likewise, the new types of
high-performance concrete usually display brittle fracture behav-
ior and, thus, they require theories that are able to model such
behavior.
The formulation that we develop in this paper coincides with
that of Jenq and Shah [18] and Carpinteri et al. [19] in assuming
that diagonal tension failure is caused by a propagation of exural
cracks. Beam failure occurs when a exural crack reaches a certain
depth, which we call critical depth [28]. It depends on the crack po-
sition and on the boundary and loading conditions. This failure cri-
terion is based on experimental observations [12] and on results
obtained with analytical models [19]. In this regard, it has been ob-
served that cracks initially progress in a stable manner [19]. When
one of them reaches its critical depth it becomes unstable, which
Nomenclature
a non-dimensional horizontal distance from the support
to the crack tip
a
0
non-dimensional initial crack mouth position
b
H
Hillerborg brittleness number
g non-dimensional bond
c non-dimensional compression force depth
k shear span slenderness ratio
l trajectory exponent
x
c
crack opening at the reinforcement level
x

c
non-dimensional crack opening at reinforcement level
r
s
reinforcement stress
r

s
non-dimensional reinforcement stress
r
y
minimum between the yielding and sliding stress for
the bars
s
c
bond strength
n non-dimensional crack depth
n
crit
non-dimensional critical crack depth
f non-dimensional reinforcement cover
A
s
bars cross section area
b beam width
c reinforcement cover
E
c
concrete Youngs modulus
E
s
steel Youngs modulus
f
c
concrete compression strength
f
ct
concrete tensile strength
G
F
concrete fracture energy
h beam height
l shear span
l
ch
concrete characteristic length
M
c
bending moment during crack progress
M

c
non-dimensional bending moment during crack pro-
gress
P reinforcement reaction
p
e
bar perimeter
T
c
traction at crack front
T

c
non-dimensional traction at crack front
T
s
bars reaction
T

s
non-dimensional bars reaction
V applied shear force
V
c
shear for concrete crack propagation
V

c
non-dimensional shear for concrete crack propagation
V
F
shear for element failure
V
t
shear for crack propagation
V

t
non-dimensional shear for crack propagation
x crack tip horizontal position
x
0
crack mouth horizontal position
y compression force depth
q reinforcement ratio
c non-dimensional depth of the compression force
y depth of the compression force
46 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556
means that the crack length increases thereby consuming energy
stored solely in the specimen with no need for additional external
energy to propagate the crack. As the model is based on Fracture
Mechanics concepts, it reproduces the size-effect observed in the
experiments and the inuence of the variables that govern the fail-
ure. It also introduces the effect of the bond between the reinforc-
ing bars and the concrete matrix on shear behavior, which proves
to be inuential in allowing or restraining crack propagation.
The papers is structured as follows. The subsequent section de-
scribes the assumptions made as to model shear crack propagation
both in plain and reinforced concrete. Section 3 deals with the fail-
ure criterion that is adopted in the model. A discussion on the
shape of the crack is included in Section 4, followed by a detailed
description of the response of the model plus its validation against
experimental results, in Section 5. Finally, Section 6 summarizes
the results of the paper and draws several conclusions.
2. Modeling shear crack propagation
2.1. Plain concrete
A three point bending beam (TPB) is deemed to exist where a
vertical crack develops at a point in the shear span. The different
geometric variables relevant to the problem are displayed in
Figs. 1a and 2a. The beam has a depth h, a width b and a shear span
equal to l, (which is the horizontal distance between the load point
and the closest support). The depth of the tensile stresses resultant
is represented as z and the depth of the compressive stresses resul-
tant as y. All these dimensions can be expressed in a non-dimen-
sional way by dividing by the depth h. In this manner, we dene
n
z
h
as the depth of the tensile force expressed in a non-dimen-
sional form and c
y
h
as the depth of the compressive force ex-
pressed in a non-dimensional form; these parameters have a
value between 0 and 1. We will also consider two additional
parameters, the shear span slenderness, which is dened as k
l
h
,
and another that indicates the distance of the crack to the beam
support, a
x
l
, where x is the horizontal distance from the crack
to the support.
The depth of the compressive force, y, is evaluated by the equi-
librium of horizontal forces, see Fig. 1a:
T
c
C
c
; 1
We assume a rectangular distribution for the compressive stresses.
The compressive force is also considered to be situated at the cen-
troid of the distribution of compressive stresses so (1) can be
rewritten as:
T
c
2bh yf
c
: 2
From Eq. (2) the depth of the compressive force is expressed as:
y h
T
c
2bf
c
: 3
Tensile and compressive forces can be expressed in a non-dimen-
sional form by dividing them by the cross section area, bh, multi-
plied by the concrete tensile strength, f
ct
.
T

c

T
c
bhf
ct
; 4
C

c

C
c
bhf
ct
: 5
Thus, the result from expressing Eq. (3) in a non-dimensional form
is:
c 1
1
2n
f
T

c
; 6
where n
f
is the ratio between the concrete compression strength, f
c
and the concrete tensile strength, f
ct
.
To represent the different non-linear processes located at the
front of the crack, the model assumes a cohesive behavior of the
fractured concrete. During the crack development a damage zone
is generated at the front of the crack, along which tensile stresses
are transferred between the crack faces during the crack opening.
The main consequence of the existence of a non-linear area at
the crack front is the Fracture Mechanics Size Effect, which is
due to the release of the stored energy of the concrete element into
the fracture zone. Thus, in this work it is considered that the Frac-
ture Mechanics size effect is the most important source of strength
scaling, as it is proposed in reference [29]. The force needed for
crack growth at a given crack depth and accounting for non-linear
effects is represented through Bazants law [30], which is well-
known and widely-disseminated among researchers in concrete
fracture mechanics. Thus, it is considered that T

c
follows Bazants
law, since it is a non-dimensional nominal strength that incorpo-
rates the cohesive stresses in the crack. Therefore, it can be written
as:
T

c

B

1
b
H
b
0
_ ; 7
where b
H
is the Hillerborgs brittleness number [29], which is de-
ned as the ratio between the depth of the beam, h, and the mate-
(a)
(b)
Fig. 1. Bending analysis: (a) plain concrete section and (b) reinforced concrete
section.
(a)
(b)
Fig. 2. Beam geometry: (a) plain concrete and (b) reinforced concrete.
J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556 47
rial characteristic length,
ch
, which in turn is dened as
ch

Ec G
F
f
2
ct
,
where E
c
is the elastic modulus of concrete and G
F
is its fracture en-
ergy. b
0
is a constant related to the aggregate size and the concrete
characteristic length [31]. B is a non-dimensional constant related
to the geometry of the element. Note that the tensile force is inde-
pendent of the shape of the cohesive function because it has been
formulated through Bazants law. The point where the tensile force
is located will be considered the reference point during the crack
growth, z, see Fig. 1a. Thus, for the sake of simplicity, z is identied
with the crack depth. It should be considered that in a cohesive or
ctitious crack approach, the position of the crack tip is smeared
along the crack process zone [32].
B in Eq. (7) represents the relationship between the reference
tensile strength and the strength for a zero-size specimen of a gi-
ven geometry [33] and for that reason it can be derived from an
asymptotic analysis [34]. Thus, the value of the non-dimensional
tensile force for a zero-size specimen (that is, for h !0 and, thus,
b
H
!0) is evaluated from Eq. (7):
lim
h!0
T

c
T

c0
B: 8
In the present work, it is assumed that the tensile strength is
reached at every point along the crack ligament for zero-size
beams, due to the fact that the crack openings for h 0 are very
small and there will not be any tensile softening. Thus, for a
zero-size beam the tensile force in the section can be expressed as:
lim
h!0
T
c
T
c0
by zf
ct
; 9
where it is also assumed that the length of ligament is the distance
between the compressive and the tensile resultants. The previous
expression can be written in a non-dimensional manner as:
T

c0
c n: 10
Identifying Eq. (8) with Eq. (10) nally allows us to write the size-
effect law for the non-dimensional tensile resultant as:
T

c

c n

1
b
H
b
0
_ : 11
The bending moment for a certain crack depth can be evaluated by
multiplying the force, T
c
, by the distance between the compressive
and tensile resultants, (y z), as shown in Fig. 1a. The resulting
non-dimensional expression is:
M

c

M
c
bh
2
f
ct
T

c
c n
c n
2

1
b
H
b
0
_ : 12
For a three point bending beam (TPB), the bending moment is also
equal to the reaction at the support multiplied by the distance to
the crack, see Fig. 2a. As the reaction at the support is equal to
the shear in a TPB beam, we can state:
M
c
Vx Vakh )M

c
V

c
ak: 13
By substituting Eq. (13) in (12), we obtain the shear for a given crack
depth, which is expressed as:
V

c

V
c
bhf
ct

1
ak
c n
2

1
b
H
b
0
_ : 14
This size-effect law is validated against the experimental results
on plain notched beams of several sizes performed by Carmona
et al. [12] as shown in Fig. 3a, where the x-axis represents the size
in terms of the Hillerborgs brittleness number and the y-axis, the
non-dimensional concrete shear for crack propagation, V

c
. The pro-
posed model produces a solid line in Fig. 3a, which reproduces the
size effect of the experimental results.
The model response for several types of concrete is shown in
Fig. 3b. In this gure the x-axis represents the non-dimensional
depth, n, and the y-axis the non-dimensional shear strength given
by Eq. (14) for plain concrete. As expected, shear strength reduces
as crack depth increases. Fig. 3b also shows that non-dimensional
shear decreases as characteristic length decreases. This is another
way of looking at size effect, since when characteristic length de-
creases the beam size, in terms of the Hillerborgs brittleness num-
ber, b
H
, increases, which means that non-linear effects have less
importance for larger sizes.
2.2. Reinforced concrete
Concrete elements are usually reinforced by steel bars posi-
tioned to resist tension. The reinforcement introduces a new force,
T
s
, in our section, as shown in Fig. 1b. The value of this force, T
s
, is
equal to the steel area, A
s
, multiplied by the tension in the bars, r
s
,
Eq. (15). The force in the reinforcement can be expressed in a non-
dimensional manner by dividing it by the section area and the con-
crete tensile strength:
T
s
A
s
r
s
)T

s

T
s
bhf
ct

A
s
r
s
bhf
ct
qr

s
; 15
where q is the reinforcement ratio and r

s
is the tension in the rein-
forcement expressed in non-dimensional form. The position of the
reinforcement is dened by the concrete cover, c, which can also
be expressed in a non-dimensional manner as a ratio of the beam
depth, f
c
h
. The different geometric variables relevant to the prob-
lem are displayed in Figs. 1b and 2b.
The value of the bending moment for a given crack depth z can
be written as:
M
t
M
c
M
s
T
c
y z T
s
y c: 16
(a) (b)
Fig. 3. Plain concrete: (a) Model validation. (b) Non-dimensional shear for crack propagation.
48 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556
Re-writing Eq. (16) in a non-dimensional form:
M

t
M

c
M

s

c n
2

1
b
H
b
0
_ qr

s
c f: 17
Since the reinforcement introduces a new force in the cross section,
the equation for the horizontal equilibrium of forces is:
C
c
T
c
T
s
: 18
The non-dimensional depth for the compressive force is now:
c 1
1
2n
f
T

c
qr

s
_ _
: 19
Finally, using Eq. (13), the shear that leads to equilibriumfor a given
crack depth can be expressed as:
V

t

V
t
bhf
ct

1
ak
c n
2

1
b
H
b
0
_ qr

s
c f
_

_
_

_: 20
Note that Eq. (20) indicates that the shear borne by the beam is the
sum of two terms. The rst one depends on the concrete properties,
whereas the other one depends on the reinforcement ratio, the ten-
sion in the steel bars and the concrete cover. The sum of these terms
is multiplied by another term which takes into account the slender-
ness of the beam and the position of the crack front. All of the values
in (20) are known except for the steel tension.
In order to evaluate the steel tension during crack growth, we
need to consider an additional equation by enforcing the compat-
ibility of displacements between concrete and steel, which implies
that the crack opening is equal to the stretching of the reinforce-
ment bars. It is written as:
x
c
2
Dl; 21
where x
c
is the crack opening and Dl is the stretching of the steel
bar. It is assumed that the traction force of the reinforcement is
equal on both sides of the crack, as shown in Fig. 4 and Eq. (21).
Other compatibility equations could be considered, for example,
based on the Navier hypothesis or depending on the crack opening,
as done in [25].
Eq. (21) can be expressed in a non-dimensional way by dividing
both terms by the depth of the beam:
x
c
2h

Dl
h
)
x

c
2
Dl

: 22
The non-dimensional crack opening, x

c
, can be evaluated by the
expression given by Tada et al. [35]. In order to take into account
the concrete cover, an additional term, 1
f
n
, has been introduced:
x

c
2

x
c
2h
12 akV

t
f
ct
E
c
nf n 1
f
n
_ _
; 23
where f n is:
0:76 2:28n 3:87n
2
2:04n
3

0:66
1 n
2
: 24
The stretching of the bar (Fig. 4) can be expressed as:
Dl

Dl
h

r
2
s
A
s
2s
c
E
s
p
e
h
r

s
_ _
2 f
2
ct
2s
c
E
s
A
s
p
e
h
; 25
where s
c
is the bond strength between steel and concrete, which is
considered constant along the adherence length, and p
e
is the bar
perimeter. By substituting Eqs. (23) and (25) in Eq. (22), the non-
dimensional tension on reinforcement can be expressed as:
r

s
_ _
2
24V

t
akg
2
b
H
nf n 1
f
n
_ _
; 26
where g is the non-dimensional bond strength dened by Ruiz [25],
which can be written as:
g

n
E
s
c
f
ct
p
e

ch
A
s

; 27
where n
E
is the ratio between the elastic modulus of steel and that
of concrete. Note that g depends not only on s
c
but also on the ratio
p
e

ch
=A
s
, i.e. g is inversely proportional to the square root of the bar
diameter,

/
p
. Thus, strong bonds can be achieved by using thin bars
(or bers) with a regular bond strength.
For the sake of simplicity, elasticperfectly-plastic behavior for
the steel of the reinforcement bars is chosen in this work. Thus,
once the tension in the reinforcement reaches the yield strength,
f
y
, it remains constant during crack propagation. Tension in the
reinforcement bars will then be given by:
r

s

24V

t
akg
2
b
H
nf n 1
f
n
_ _ _ _
1=2
if . . .
1=2
<
fy
fct
f
y
f
ct
otherwise:
_

_
28
In order to evaluate shear force and tension in steel bars, we nally
have a system of three equations, Eqs. (19), (20) and (28), which can
be solved analytically. When the reinforcement traction reaches the
yield strength, the shear strength is obtained just from Eqs. (19) and
(20). It should be noted that the proposed model is valid for low and
medium reinforcement ratios. For high ratios, the failure is caused
by excessive compressions below the load bearing point, and study-
ing this type of failure is beyond the scope of this paper. Finally, we
assume that a crack may form at any point along the shear span,
which is especially true in the case of ribbed bars.
In Fig. 5bd, the model response is shown for the midspan sec-
tion of the beam in Fig. 5a. A beam under three point bending is
modelled. Fig. 5a also shows material properties. It is assumed that
cracks grow in a vertical manner. In Fig. 5b the x-axis represents
the non-dimensional depth, n, and the y-axis, the non-dimensional
shear strength during crack growth, V

t
. A curve is determined for
each reinforcement ratio. As the reinforcement ratio increases,
the shear for steel yielding increases.
It is well-known that crack growth may present stable or unsta-
ble behavior. When crack growth is stable, an increase in crack
depth requires a load increase. Conversely, unstable crack growth
is accompanied by a load decrease. If we observe the curve for
q 0:78% in Fig. 5b, an unstable branch is observed after crack ini-
tiation. When the crack reaches the reinforcement an increase in
the load is detected and, afterwards, another unstable branch oc-
curs. There exists a minimum for V

t
at n 0:6 beyond which crack
growth becomes stable. From this point on, shear increases until
the reinforcement yields and the exural capacity of this beam sec-
tion is reached. The behavior obtained with our model is analogous
to the one described by Carpinteri using the shear version of his Fig. 4. Compatibility of displacements.
J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556 49
Bridged Crack Model [36]. Nevertheless, the model proposed intro-
duces concepts of cohesive theories in order to analyze size effect
and it allows studying the inuence of bond between concrete and
steel during the crack growth.
In Fig. 5c and d, the x-axis represents the non-dimensional crack
opening at the reinforcement level, x, and the y-axis corresponds
to the non-dimensional shear during crack growth, V

t
. As the rein-
forcement ratio increases, the shear also increases for the same
crack width. Fig. 5d zooms in on Fig. 5c for low values of the crack
width. In the case of plain concrete (q 0), it can be observed that
shear rapidly decreases to zero and that cracks never advance in a
stable manner.
The methodology presented here analyzes shear load during the
propagation of cracks in the case of three point bending beams.
Beams subject to uniformly distributed loads and other load and
boundary conditions, may be studied with this methodology there-
by changing the equilibrium equations in order to achieve a new
relationship between shear and applied load. Moreover, the model,
conceptually simple as it is, shows that the equilibrium shear for a
given crack depth and in a given position along the shear span is
governed by nine non-dimensional parameters. These parameters
can divided into three categories.
Geometrical parameters:
n
z
h
; k
l
h
; a
0

x
l
; f
c
h
: 29
Parameters referring to size and concrete properties:
b
H

h

ch
; n
f

f
c
f
ct
: 30
Parameters referring to steel properties and bond between steel and
concrete:
q
A
s
A
c
; g

n
E
s
c
f
ct
p
e

ch
A
s

; f

y

f
y
f
ct
: 31
Eq. (20) demonstrates that these three parameters are com-
bined through the product qr

s
. Before yielding, the response of
the model depends on the product qg, whereas it is qf

y
that gov-
erns the second term in Eq. (20) after the reinforcement yields.
3. Failure criterion
Eq. (20) calculates the shear force for crack propagation, V
t
, but
shear failure, V
F
, will occur only for a denite crack depth. This
depth will be referred to henceforth as critical depth, n
crit
. A crite-
rion to determine n
crit
can be derived from experimental observa-
tions, namely those made by Carmona et al. [12]. This
experimental program was designed so that only one single
mixed-mode crack generated and propagated through the speci-
men, as opposed to the usual dense crack pattern found in most
of the tests reported in the literature. In Fig. 6, we illustrate two
of the results that will help to explain the failure criterion.
Fig. 6a shows the crack pattern in two of the tests. The marks
and gures on the sketch refer to the corresponding points in the
Pd curves, as shown in Fig. 6b, and to the load in kN that the beam
was resisting when the crack tip reached that position. During the
crack progress, a change in the nature of the crack propagation was
observed and a subsequent unstable crack branch began leading to
the beam failure. This phenomenon was associated with the so-
called diagonal tension failure. This change in the nature of crack
propagation can be observed in point C of beam L40 and point D
of beam L80. These points shown in Fig. 6a are approximately lo-
cated on the line that joins the loading point with the point where
the reinforcement reaches the support. This line coincides with an
5000mm
(a)
Materials properties
Concrete
f (MPa) c f (MPa) t G (N/m)
F
E (GPa)
c
30.0 3.2 120.0 33.0
Ribbed steel
f (MPa)
y 0,2%
(MPa)
c
E (GPa)
c
520 200
2.0
200
Dimensions in mm
225 220 212
) % 2 2 . 1 ( ) % 0 . 0 ( (0.78%) (0.28%)
400 350
Reinforcement
ratio
Notation
cross section
0
0.1
0.2
0.3
0 0.2 0.4 0.6 0.8 1.0
=0.00%
=0.28%
=0.78%
=1.22%

V
*
t
0
0.1
0.2
0.3
0 0.02 0.04 0.06 0.08 0.1

V
*
t
0
0.025
0.05
0 0.0005 0.001
V
*
t
(b)
(c)
(d)
=0.00%
=0.28%
=0.78%
=1.22%
=0.00%
=0.28%
=0.78%
=1.22%
c
*
c
*
Fig. 5. Model response: (a) geometry and properties of the materials; (b) V

t
n curves; (c) V

t
x

c
curves; and (d) detail Vt x

c
curves.
50 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556
ideal strut that connects the loading point to the support. These
experimental observations were also assumed in an analytical
model proposed by Carpinteri et al. [19]. Therefore, for three point
bending exure, the critical depth is dened by the line connecting
the loading point to the point where the reinforcement reaches the
support. This denition was previously proposed by Zararis and
Papadakis [28]. Mathematically, it can be expressed as follows:
n
crit
f a1 f: 32
This failure criterion is associated with three point bending exure,
but can be easily generalized for any boundary and load conditions.
For example, for a simply-supported beam subjected to a uni-
formly-distributed load, the bending moment diagram variation is
parabolic with a maximum at midspan. The critical crack variation
is also parabolic. At the maximum bending moment point, the crit-
ical crack is equal to the beam depth, and in the support it is equal
to the cover; see Fig. 7b. Therefore, we propose that the critical
depth be related to the bending moment diagram: For the maxi-
mum bending moment position, the critical depth is equal to the
beam depth (n
crit
1), and when the bending moment is equal to
zero, the critical depth is equal to the reinforcement concrete cover.
In between, the critical depth is proportional to the value of the
bending moment at that position.
4. Crack shape
So far we have assumed that cracks grow in a vertical manner
from the initiation point. In reality, cracks have curved trajectories
approximately following the lines indicating the direction of max-
imum compression. This effect means that the initial abscissa of
the crack, a
0
, is not the same as the abscissa of the crack tip when
it reaches critical depth, a. In order to consider this mismatch,
L80-9-2
14
15.2
15.7
16
15.5
15
12
12
15
15.6
15.6
15
14.5
14
13
16.4
D
B
E
B
A
C
C
D
0
4
8
12
16
0 1.5 3
(mm)
P

(
k
N
)
L80-9-2
F
E
D
B
A
C
(b)
(a)
L40-7-2
9.4
9.5
8.6
8.1
9.85
9.8
9.7
9.5
A
D
C
B
0
2
4
6
8
10
12
0 1 2
(mm)
P

(
k
N
)
L40-7-2
D
C
B
A
1
15
16
1
D
E
B
CC
8
9
9.8
DD
CC
B
failure
failure
failure
failure
Fig. 6. Failure criterion, experimental results: (a) crack patterns and (b) Pd curves.

crit

crit
(a)
(b)
P
q
Fig. 7. Failure criterion: (a) three point bending and (b) uniformly distributed load.

crit
0
Fig. 8. Crack shape.
J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556 51
equations dening trajectory patterns can be used, as the ones
shown below, which were proposed in reference [19]:
af; n
a
0
0 6 n 6 f
a
0

nf
1f
_ _
l
1 a
0
f 6 n 6 1:
_
33
The above formula assumes a straight trajectory from the initiation
point to the reinforcement and a parabolic trajectory which reaches
the load-bearing point. Through Eq. (33), a relationship between the
crack depth n and the initiation point, a
0
, is determined. For three
point bending, exponent l is equal to:
l
1
1 a
0
: 34
This equation was obtained from the experimental tests performed
by Carpinteri et al. [37]. The failure criteria for curved cracks remain
the same as for straight cracks: Once the critical depth is reached,
the element fails, (see Fig. 8).
5. Results and discussion
5.1. Model response and experimental validation
In this section it will be shown how the value of the initial crack
position, a
0
, affects the shear strength. A beam under three point
bending is modelled. Fig. 9a shows the geometry and material
properties. In Fig. 9b the x-axis represents the non-dimensional
depth, n, and the y-axis the nondimensional shear strength during
crack growth, V

t
. A curve is determined for each crack initiation
position. The black part of the curve describes the shear resistance
variation during the crack process until the critical depth is
reached (see Fig. 9b). The point where the crack front reaches the
critical depth indicates shear strength.
As it can be observed in Fig. 9b, failure occurs prior to reinforce-
ment yielding (brittle failure, diagonal tension) for cracks which
form close to the support, whereas failure occurs after the rein-
forcement yields (exural failure) for the ones farther off the sup-
port. It must be noted that once reinforcement has yielded there
can also be brittle failure, as shown by Rodrigues et al. [38]. Never-
theless, when the crack is located under the loading point, as the
critical depth is equal to the depth, a brittle failure caused by diag-
onal tension cannot occur.
To validate the response of the model, we have compared the
results obtained with it to those obtained by a recent experimental
program performed by Carpinteri et al. [37]. A total of sixteen geo-
metrically similar beams reinforced with 4 different reinforcement
ratios were tested (4 beams for each ratio), see Fig. 10a. In this
experimental program, the initial position and the shape of critical
cracks were studied. The residual bond strength between concrete
and steel was not measured in the experimental program and,
therefore, it has been estimated using the formulation established
in the Model Code (CEB-FIB), see Fig. 10a. Fig. 10b shows the crit-
ical crack depth considered and the notation in order to facilitate
the understanding of the plots in Fig. 10c.
Fig. 10c shows the comparison between the theoretical and
experimental results. The x-axis corresponds to the initial position
of the cracks, whereas the y-axis represents the shear when the
crack reaches the critical depth (shear strength), V

F
. In order to
facilitate the comparison, selected crack patterns corresponding
to one of each kind of the tested specimens are drawn at the bot-
tom of the gure. The continuous lines represent the models re-
sponse while the symbols represent the experimental results.
In order to describe the models response, we are going to focus
on the curve corresponding to a beam reinforced with 2/12, start-
ing from the section corresponding to the load application point
(a
0
1). For the crack under the loading point, the rebar yields be-
fore the crack reaches the critical depth and the failure is due to
exure. Distancing the initiation of the crack from the loading
point, we observe that the shear strength increases, since the rein-
forcement gets more load and may even yield before the critical
depth is attained. For a certain point, a
Y
, a maximum value on
the curve is detected. This maximum indicates the point at which
the rebar reaches its yield strength exactly at the same time that
the crack reaches its critical depth. For values of a
0
lower than
a
Y
, the rebar has not yielded when the crack tip reaches the critical
depth and the beam failure occurs by diagonal tension. The shear
strength decreases until reaching a minimum for a certain initia-
tion point, situated in our case at a
0
0:4. This minimum has also
been found in experimental results performed by Kim and White
[39,40]. For cracks with an initiation point closer to the support,
the shear strength starts to increase, although the critical depth
is low. It should be noted that the actual shear strength of the beam
is the smallest shear found by varying a
0
.
The shape of the curve obtained with the model, demonstrating
a minimum in the central part of the shear span, coincides with the
description proposed by Kani and Wittkopp [20]. In the zone near
the support, we nd an area where the failure is produced due to
the yielding of the rebars (exure); but as we move away from
the support along the shear span, the failure occurs during the
development of a crack (diagonal tension).

crit
1400mm
1200mm
100
Dimensions in mm.
28
(0.50%)
200 180
Reinforcement
ratio
Notation
(a)
Materials properties
Concrete
f (MPa)
c f (MPa)
t
G (N/m)
F
E (GPa)
c
49.4 3.2 111.5 33.1
Ribbed steel
f (MPa)
y 0,2%
(MPa)
c
E (GPa)
c
519 200
1.1
0
1
0
0.5
1.0
1.5
0 0.2 0.4 0.6 0.8 1.0
0=1.0
0=0.8
0=0.6
0=0.4
0=0.2
V = V ( )
crit F
*
V
*
t
(b)
1.0 0

t
*
Fig. 9. Model response: (a) geometry and properties of the materials and (b) V

t
n
curves.
52 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556
Fig. 10c shows that for low reinforcement ratios, e.g. 1/8, the
failure takes place in sections near the loading point, where the
steel reaches the elastic limit before the crack grows to the critical
depth. Upon increasing the reinforcement ratio, the critical section
moves away from the point of application of the load to positions
where the steel does not yield. The experimental tendency is cap-
tured by the model and even the shear loads obtained coincide
quite reasonably with the experimental ones. Moreover, the model
explains the valley in the shear resistance along the beams lon-
gitudinal axis. In the results obtained for the largest reinforcement
ratio (2/20), the differences between the model and the experi-
ments are conditioned by the type of failure, since in highly-rein-
forced beams the failure is produced by excessive compression
and not by diagonal tension.
5.2. Size effect
The sensitivity of the models response to size is validated
against experimental results on notched reinforced concrete beams
of several sizes performed by Carmona et al. [12], Fig. 11. The x-
axis represents the size in terms of the Hillerborgs brittleness
number and the y-axis, the non-dimensional concrete shear for
crack propagation, V

c
. The model follows a trend similar to that
of the tests analyzed.
Fig. 12 shows the models response when varying the element
size. Fig. 12a displays the size effect when the reinforcement ratio
is increased. The x-axis represents the size in terms of the Hiller-
borgs brittleness number and the y-axis, the non-dimensional
shear strength, V

t
. A curve is determined for each reinforcement
ratio. When there is no reinforcement, the curve follows Bazants
law. As the reinforcement ratio is increased, the size effect changes.
For a given reinforcement, the curve is similar to that of plain con-
crete only for small values of b
H
. As size increases, the contribution
of the reinforcement grows. A minimum in the non-dimensional
shear strength is reached and, henceforth V

F
grows with size until
*
V
F
18
28
212
220
1400mm
1200mm
100
Dimensions in mm
18 220 212 28
) % 4 1 . 3 ( ) % 5 2 . 0 ( (1.13%) (0.50%)
200 180
Reinforcement
ratio
Notation
(a)
0
0.5
1
1.5

0
tests
model results
18
28
212
220
Diagonal tension
failure Flexural failure
18-3
212-2
28-3
220-2
llll
(c)
P

crit
0 0.25 0.5 0.75 1

0
Materials properties
Concrete
f (MPa)
c f (MPa)
t
G (N/m)
F
E (GPa)
c
49.4 3.2 111.5 33.1
Ribbed steel
f (MPa)
y 0,2%
(MPa)
c
E (GPa)
c
519 200
1.1
(b)
0 0.25 0.5 1 0.75
Fig. 10. Model response: (a) beam geometry and (b) experimental results vs. model, V

F
a0 curves.
=0.026%
test =0.026%
test =0.013% =0.013%
=2 0=0.5 n =9.5
f
f =148
y
*
=15
V
*
F
0.05
0.1
0.15
0.2
0.5 1 2

H
5
Fig. 11. Size effect: experimental results vs. model.
J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556 53
the reinforcement bars yield. Of course, afterwards V

F
becomes
insensitive to size. A similar trend in size effect for reinforced con-
crete beams was obtained by Ruiz [25]
Fig. 12b compares the response for two different values of the
crack initiation position: The rst crack starts at the middle of
shear span, a
0
0:5 (diagonal tension failure), and the other initi-
ates under the load point, a
0
1:0 (exural failure). The x-axis rep-
resents the element size in terms of Hillerborgs brittleness
number. The usual range of this parameter in structures is indi-
cated with two vertical lines. The y-axis represents the non-dimen-
sional shear strength. For the crack located at a
0
0:5, there exists
a strong size effect, i.e. shear strength depends on element size,
which is caused by the existence of a fracture process zone on
the crack front. Nevertheless, size effect disappears both for very
small sizes as well as for large ones. This can be explained with
the ratio between the length of the damage zone generated at
the front of the crack and the beam height. Recall that the length
of the damage zone is a material characteristic and thus, its value
tends to be constant for a given concrete as size increases, i.e. it
tends to be very short compared with the beam depth for large
beams. In other words, the rst term in Eq. (20) tends to be 0 as
size increases, and then the equilibrium shear is given only by
the second term. This second term gives a constant value for large
elements because the reinforcement bars yield and thus the nom-
inal stress is constant and size independent. On the other hand, for
small beams the damage zone tends to occupy the entire beam
height and size effect is greatly reduced. Therefore, the rst term
in Eq. (20) tends to a constant value as size decreases.
For the crack located at midspan, a
0
1:0, shear strength does
not present any size effect, since the term (c n
crit
) in Eq. (20) is
equal to zero. This means that non-linear stresses at the crack front
do not contribute to the equilibrium shear and thus, size effect dis-
appears. Moreover, for a completely-developed crack the rebars
yield and so the stress is constant and independent of the size.
The results obtained with the model coincide with the experi-
mental observations made by Collins and Kuchma [6]. They
showed that the size effect is greatly reduced in beams containing
well-distributed longitudinal reinforcement along the beam
(a) (b)
0=0.5
0=1.0
usual structural
dimensions range
=3
=5 n =10
f
f =125
y
*
=0.5%
0.1 1 10 100

H
=0.0%
=0.1%
=0.25%
=0.5%
=1.0%

H
0.001 0.01 0.1 1 10 100 1000
V
*
F
=3 0=0,5
0.001
0.01
0.1
1
0.01
0.1
1
V
*
F
Fig. 12. Size effect: (a) inuence of reinforcement ratio and (b) inuence of crack initiation position.
(a)
(b)
0
0.2
0.4
0.6
0 0.2 0.4 0.6 0.8 1

V
*
t
(c)
0
0.5
1.0
1.5
0 0.25 0.5 0.75 1.0

0
V
*
F
(d)
100
112
200
0
0.2
0.4
0.6
0 0.0025 0.005 0.0075 0.01
V
*
t
0
0.0025
0.005
0.0075
0,01
0 0.2 0.4 0.6 0.8 1

=5
=15
=50
=150
=5
=15
=50
=150
=5
=15
=50
=150
=33
=50
=10
=150

c
*

c
*
Fig. 13. Bond inuence: (a) V

t
n curves; (b) V

t
x

c
curves; (c) x

c
n curves; and (d) V

F
a0 curves.
54 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556
height. The presence of reinforcement avoids the unstable crack
propagation and thus changes the type of failure from diagonal
tension to exure, where size effect is less noticeable.
5.3. Bond effect
The inuence of bond between concrete and reinforcing bars is
introduced in the model (Eq. 28) using the non-dimensional bond
strength, g (Eq. 27), dened by Ruiz [25]. As already mentioned,
this parameter depends not only on s
c
, but on the ratio p
e

ch
=A
s
,
i.e. g is inversely proportional to the square root of the bar diame-
ter,

/
p
. This section addresses how bond characteristics affect
shear strength and size effect.
Fig. 13 shows the model response for variations in the bond
characteristics between concrete and steel. The assumed range
for g is set to visualize the bond effect, since the study of asymp-
totic or extreme cases helps to understand the inuence of bond
in the behavior of reinforced concrete elements.
Fig. 13a represents crack depth versus shear. We modelled the
beam in Fig. 12a reinforced with 1/12. All the parameters were
kept constant except for the bond strength. When bond is in-
creased, the crack depth at which the steel yields decreases. For
low bond strength conditions (smooth bars), cracks have to devel-
op fully before the steel yields. In the case of high bond strength
(ribbed bars), steel yielding occurs shortly after the crack crosses
the rebar. In Fig. 13b, the x-axis displays the crack width and the
y-axis, the shear strength. As the bond strength increases the
behavior of the reinforced concrete section is stiffer, since steel
yielding is reached with shorter crack widths. Fig. 13c represents
crack depth versus crack width. It reveals that bond strength con-
trols crack width. Stronger interfaces lead to wider openings if
compared to the same crack length. Of course, in this case the sec-
tion with stronger bond strength resists considerably more shear.
In Fig. 13d, the x-axis displays the initial crack position and the
y-axis represents the shear strength. As bond strength increases,
the value of a
Y
decreases and the shear strength increases. In the
extreme case that the bond strength has an innite value, the ele-
ment fails by exure because, for all possible cracks, the steel
yields just after crossing the rebar position and therefore, before
reaching the critical depth. In this case, the failure takes place at
the section under load bearing point (a
0
1:0)
Fig. 13b also shows that the minimum of the curve moves to-
wards the support as the bond strength increases. Low values for
the bond strength lead to low values for the failure loads, even
lower than the load that is required for fracture initiation at that
particular point, V
fis
. This means that the diagonal tension failure
would be very unlikely for beams in which the steel-to-concrete
interface is weak. Indeed, in this case, there would be cracks only
under the load-bearing point and there would be neither crack
generation nor propagation along the shear span. Therefore, in
such conditions, shear failure would not occur.
Fig. 14 shows the inuence of non-dimensional bond on size ef-
fect. The mechanical properties in these examples are shown in the
graphic. The x-axis represents the size in terms of the Hillerborgs
brittleness number and the y-axis, the nondimensional shear
strength, V

t
. As bond increases, size effects tend to disappear be-
cause reinforcement bars yield at a lower crack depth. The effect
of an increase in bond conditions is similar to the increase of rein-
forcement ratio.
6. Conclusions
In this study, a new model based on nonlinear Fracture Mechan-
ics concepts is introduced to analyze shear strength in reinforced
concrete beams without stirrups. It identies the variables that
govern the failure, including the bond between concrete and steel.
The following main points can be drawn from the study:
1. A failure criterion based on crack growth is introduced. When a
crack reaches a given depth, which we refer to as critical depth,
beam failure takes place. This critical depth depends on the type
of loading, the boundary conditions and the beam geometry.
2. Shear strength depends on the section position along the shear
span, thereby showing a minimum around the mid-shear span
(critical section). This variation is explained based on the rela-
tionship which exists between crack propagation and failure.
3. Size effect in shear strength is also reproduced by this approach,
since it stems from non-linear Fracture Mechanics concepts. It
also explains how reinforcement ratio, initial crack location
and bond between concrete and steel affect size effect and their
respective asymptotic behavior.
4. Transition between exural and shear failure is dependent on
reinforcement ratio and steel strength. As reinforcement ratio
increases the critical section moves away from the point of
application of the load to positions where the steel does not
yield.
5. The bond strength of steel-to-concrete interface is also
accounted for in the model through the parameter g which
depends not only on s
c
, but on the ratio p
e

ch
=A
s
. It has been
observed that stronger interfaces lead to narrower openings
and shorter cracks for the same applied load. Therefore, shear
failure tends to disappear as bond strength increases because
a stronger bond leads the steel bars to yield before the crack
reaches the critical depth thereby causing shear failure.
The presented model can contribute to a better understanding
of the nature of shear strength in reinforced concrete elements
without stirrups. Furthermore, it can be extended to other loading
and support conditions. Finally, expressions derived from this
study may be used to improve shear analysis of RC beams in design
codes.
Acknowledgement
Financial support from the Subdireccin General de Proyectos de
Investigacin, Spain, through Grant MAT2012-35416 is greatly
appreciated.
References
[1] Zsutty T. Beam shear strength prediction by analysis of existing data. ACI
Struct J 1968;65(11):94351.
[2] Cladera A, Mari A. Shear design procedure for reinforced normal and high-
strength concrete beams using articial neural networks. Part 1: Beams
without stirrups. Eng Struct 2004;26(7):91726.
=0.1
=1
=10
=25
=3
0=0,5
n =10
f f =125
y
*

H
=0.5%
0.001 0.01 0.1 1 10 100 1000
=50
0.01
0.1
1
V
*
F
Fig. 14. Inuence of non-dimensional bond on size effect.
J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556 55
[3] Jung S, Kim K. Knowledge-based prediction of shear strength of concrete
beams without shear reinforcement. Eng Struct 2008;30(6):151525.
[4] Schlaich J, Schafer K, Jennewein M. Toward a consistent design of structural
concrete. J Prestr Concr Inst 1987;32(3):74150.
[5] Hsu T. Unied theory of reinforced concrete. Boca Raton, FL: CRC Press; 1993.
[6] Collins MP, Kuchma D. How safe are our large, lightly reinforced concrete
beams slabs and footings? ACI Struct J 1999;99(4):48290.
[7] Vecchio F, Collins M. Predicting the response of reinforced concrete beams
subjected to shear using the modied compression eld theory. ACI Struct J
1998;85(3):25868.
[8] Muttoni A, Ruiz MF. Shear strength of members without transverse
reinforcement as a function of the critical shear crack width. ACI Struct J
2008;105(2):16372.
[9] Reinhardt HW. Similitude of brittle fracture of structural concrete. In:
Advanced mechanics of reinforced concrete, IASBE Colloquium, Delf; 1981. p.
17584.
[10] Kani G. How safe are our large concrete beams? ACI Proc J 1967;64(3):12841.
[11] Bazant ZP, Kazemi MP. Size effect in diagonal shear failure. ACI Struct J
1991;88(3):26876.
[12] Carmona JR, Ruiz G, del Viso JR. Mixed-mode crack propagation through
reinforced concrete. Eng Fract Mech 2007;74(17):2788809.
[13] Gustafsson PJ. Fracture mechanics study of non-yielding materials like
concrete: modeling of tensile fracture and applied strength analyses. Report
No. TVBM-1007, Division of Buildings Materials, Lund Institute of Technology,
Lund, Sweden; 1985.
[14] Hillerborg A, Gustafsson PJ. Sensitivity in shear strength of longitudinally
reinforced concrete beams to fracture energy of concrete. ACI Struct J
1988;85(3):2862294.
[15] Hillerborg A, Modeer M, Petersson PE. Analysis of crack formation and crack
growth in concrete by means of fracture mechanics and nite elements. Cem
Concr Res 1976;6:77382.
[16] Bazant ZP, Yu Q. Designing against size effect on shear strength of reinforced
concrete beams without stirrups: I. Formulation. J Struct Eng ASCE
2005;131(12):187785.
[17] Bazant ZP, Yu Q. Designing against size effect on shear strength of reinforced
concrete beams without stirrups: II. Verication and calibration. J Struct Eng
ASCE 2005;131(12):188697.
[18] Jenq YS, Shah SP. Shear resistance of reinforced concrete beams a fracture
mechanics approach. In: Li V, Bazant ZP, editors. Fracture mechanics:
applications to concrete. Detroit: American Concrete Institute; 1989. p.
23758.
[19] Carpinteri A, Carmona JR, Ventura G. Propagation of exural and shear cracks
through reinforced concrete beams by the bridged crack model. Mag Concr Res
2007;59(10):74356.
[20] Kani MHM, Wittkopp R. Kani on shear in reinforced concrete. Departament of
Civil Engineering, Univeristy of Toronto, ON, Canada, Toronto; 1979. 97p.
[21] Niwa J. Size effect analyses for exural strength of concrete beams using non
linear rod element. Proc JCI 1993;15(2):7580.
[22] Niwa J. Size effect in shear of concrete beams predicted by fracture mechanics.
In: CEB bulletin dInformation No. 137: concrete tension and size effects,
Comite Euro-International du Beton (CEB), Lausanne, Switzerland; 1997. p.
14758.
[23] Carpinteri A, Carmona JR, Ventura G. Failure mode transitions in RC beams.
Part 1: Theoretical model. ACI Struct J 2011;108(3):27785.
[24] Ruiz G, Elices M, Planas J. Experimental study of fracture of lightly reinforced
concrete beams. Mater Struct 1998;31:68391.
[25] Ruiz G. Propagation of a cohesive crack crossing a reinforcement layer. Int J
Fract 2001;111(3):26582.
[26] Hededal O, Kroon IB. Lightly reinforced high-strength concrete. M.Sc. Thesis,
University of alborg, Denmark; 1991.
[27] Hawkins NH, Hjorseset K. Minimum reinforcement requirement for concrete
exural members. In: Carpinteri A, editor. Application of fracture mechanics to
reinforced concrete. London: Elsevier; 1992. p. 379412.
[28] Zararis P, Papadakis G. Diagonal shear failure and size effect in RC beams
without web reinforcement. J Struct Eng ASCE 2001;127(7):73342.
[29] Bazant ZP, Planas J. Fracture and size effect in concrete and other quasibrittle
materials. Boca Raton: CRC Press; 1998.
[30] Bazant ZP. Size effect in blunt fracture: concrete, rock, metal. J Eng Mech
ASCE 1984;110:51835.
[31] Bazant ZP, Pfeiffer P. Determination of facture energy from size effect and
brittleness number. ACI Mater J 1987;84(6):46380.
[32] Petersson PE. Crack growth and development of fracture zones in plain
concrete and similar materials. Report No. TVBM-1006, Division of Building
Materials, Lund Institute of Technology, Lund, Sweden; 1981.
[33] del Viso J, Carmona JR, Ruiz G. Shape and size effects on the compressive
strength of high strength concrete. Cem Concr Res 2008(38):38695.
[34] Planas J, Guinea G, Elices E. Generalized size effect equation for quasibrittle
materials. Fatigue Fract Eng Mater Struct 1997;20(5):67187.
[35] Tada H, Paris H, Irwin G. The stress analysis of cracks handbook. Del Research
Corporation; 1973.
[36] Carpinteri A. Stability of fracturing process in RC beams. J Struct Eng ASCE
1984;110:54458.
[37] Carpinteri A, Carmona JR, Ventura G. Failure mode transitions in RC beams.
Part 2: Experimental tests. ACI Struct J 2011;108(3):28693.
[38] Rodrigues R, Muttoni A, Fernandez-Ruiz M. Inuence of shear on rotation
capacity of reinforced concrete members without shear reinforcement. ACI
Struct J 2010;107(5):51625.
[39] Kim W, White RN. Shear-critical cracking in slender reinforced concrete
beams. ACI Struct J 1999;96(5):75765.
[40] Kim W, White RN. Hypothesis for localized horizontal shearing failure
mechanism of slender RC beams. J Struct Eng ASCE 1999;125(10):112635.
56 J.R. Carmona, G. Ruiz / Engineering Structures 66 (2014) 4556