(This is a sample cover image for this issue. The actual cover is not yet available at this time.

)
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elseviers archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Behavior of reinforced concrete haunched beams subjected to cyclic
shear loading
Hans I. Archundia-Aranda
a,b
, Arturo Tena-Colunga
a,
, Alejandro Grande-Vega
a
a
Departamento de Materiales, Universidad Autnoma Metropolitana, Edicio P4, ltimo Piso, Av, San Pablo 180, Col. Reynosa Tamaulipas, 02200 Mxico, DF, Mexico
b
Divisin de Estudios de Posgrado, Facultad de Ingeniera, UNAM, Ciudad Universitaria, 04510 Mxico, DF, Mexico
a r t i c l e i n f o
Article history:
Received 18 November 2011
Revised 23 October 2012
Accepted 24 October 2012
Keywords:
Haunched beams
Tapered elements
Nonprismatic elements
Shear strength
Deformation capacity
Cyclic testing
a b s t r a c t
Final research results and interpretations of the testing of ten prototype simply-supported reinforced
concrete beams (eight haunched and two prismatic) designed to develop a shear failure under cyclic
loading are presented. Five beams (four haunched and one prismatic) were tested without shear rein-
forcement whereas the remaining ve (four haunched and one prismatic) were tested with a shear rein-
forcement satisfying the minimum required for prismatic beams by the concrete norms of Mexicos
Federal District Code (MFDC). These haunched beams are identical in geometry and reinforcement to a
set of haunched beams previously tested under static loading. These additional cyclic testing conrmed
the usefulness of a proposed equation to estimate the shear strength of reinforced concrete haunched
beams, taking into account parameters such as the haunch angle, the concrete compressive strength,
the shear reinforcement and the contribution of the inclined longitudinal reinforcement. From the
obtained results, it can be observed that haunched beams have a different cyclic shear behavior with
respect to prismatic beams, having higher deformation and energy dissipation capacities, among other
reasons, because non-prismatic beams favor an arching action in the haunched length as the main resist-
ing mechanism, that develops smoother cracking patterns.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
Reinforced concrete haunched beams (RCHBs) are often used in
simply-supported or continuous bridges worldwide (Figs. 1 and 2)
and in midrise framed buildings in Latin American countries like
Mexico and Ecuador or in European countries like Germany. Rein-
forced concrete haunched beams (RCHBs) are often used because
they offer some structural and nonstructural advantages over pris-
matic elements, such as stiffness or moment capacity to self-
weight ratio, providing an smaller effective height at midspan for
supporting frames (bridges and buildings) which eases the place-
ment of different facilities (i.e., air conditioning, piping) along the
building. Nevertheless, in some countries RCHBs are unlikely a
common structural solution in buildings because their use involve
higher construction costs, as special formwork and qualied con-
struction workers are required. In particular, haunched beams have
been used in RC bridges and buildings in Mexico City from a long
time ago. Despite this fact, there are no specic recommendations
for haunched beams in reinforced concrete norms for Mexico [1]
(NTCC-04, 2004) and the United States [2] (ACI-318-11) that would
insure the ductile detailing of these elements.
Although the German code [3] (DIN 1045-1, 2001) and some
textbooks [48] have some brief recommendations for the shear
design of RCHBs, these guidelines do not include experimental
data. The experimental evidence available is reduced to monotonic
tests of prototypes failing in shear [914]. The analytical research
related to the shear design of RCHBs is also limited. Besides the
works in which the shear resistance of a haunched section is the
sum of the contribution of the concrete, plus the contribution of in-
clined and transverse steel reinforcements [45,7,914], few works
are available where a plastic truss model has been used to analyze
the haunched element [6,8,1517].
In order to insure the desirable ductile behavior of RCHBs
according to capacity-design rules, it is necessary rst to under-
stand how sudden failures under monotonic and cyclic loads occur,
for example, the shear failure. Once this goal is achieved, it can be
possible to study how to warrant a ductile exural failure. There-
fore, in this paper the experimental results of ten RCHBs designed
to develop shear failure and tested under cyclic loading are pre-
sented. This study complements a previous study of twin RCHB
but subjected to static loading [14]. The described study is summa-
rized in following sections.
0141-0296/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.engstruct.2012.10.037

Corresponding author. Tel.: +52 55 5318 9460; fax: +52 55 5318 9085.
E-mail addresses: archundia_aranda@yahoo.com.mx (H.I. Archundia-Aranda),
atc@correo.azc.uam.mx (A. Tena-Colunga), alexgrande_ein@yahoo.com.mx
(A. Grande-Vega).
Engineering Structures 49 (2013) 2742
Contents lists available at SciVerse ScienceDirect
Engineering Structures
j our nal homepage: www. el sevi er. com/ l ocat e/ engst r uct
Author's personal copy
2. Description of test beams
The geometry of prototypes RCHBs was dened according to a
survey conducted in existing bridges and buildings in Mexico City
(Figs. 2 and 3). The width (b) for all beams was 220 mm, the effec-
tive span (L) was 2800 mm, and the shear span (a) was 1083 mm.
The haunched length (L
h
) at both beam ends was one-third the
effective span (L
h
= L/3 933 mm). Five different linear tapering
geometries were obtained by keeping constant the overall depth
at each beam end (h
max
= 450 mm) and reducing the overall depth
at the central prismatic length to h
min
= 450 (prismatic control ele-
ment), 400, 350, 300 and 250 mm. Therefore, haunched angles
from the horizontal (a) were 0, 3.07, 6.12, 9.13 and 12.10
respectively.
The geometry of all prototypes satised the requirement L/h > 5
to be considered as slender beams by the Mexican code (L/
h
max
> 5). In addition, with the purpose of not magnifying the char-
acteristic arching mechanism observed experimentally and analyt-
ically in haunched beams [9,10,18], all prototypes were checked to
fulll the well-known a/d limiting ratio between slender beams
and short beams (a/d
max
> 2.5). The top and bottom reinforcement
cover was 40 mm.
The specied material properties for design were a compressive
strength f
0
c
24:5 MPaf
0
c
250 kg=cm
2
for the concrete, and a
yield tensile stress f
y
= 412 MPa (f
y
= 4200 kg/cm
2
) for all the steel
reinforcement. The measured properties for the steel reinforce-
ment from coupon tests are given in Table 1. The measured prop-
erties from compression tests of small cylindrical specimens at the
date of testing are identied in Table 2.
Two beams were constructed for each one of the ve different
geometries considered: (a) one beam without shear reinforcement,
only with four stirrups outside the shear span to hold the longitu-
dinal steel reinforcement and (b) one beam with minimum shear
reinforcement according to NTCC-04 for prismatic elements con-
sidering the effective depth at the support (d
max
). Therefore, ten
prototype haunched beams were tested.
To insure a shear failure along the haunches, the design was
made providing the exural capacity in the central prismatic
length and keeping continuous the longitudinal reinforcement
along the prototypes. As described in a previous paper [14], all
the specimens (monotonic and cyclic) were originally designed to
insure that they failed in shear while following general NTCC-04
guidelines for prismatic beams. The following considerations were
done:
(a) The nominal contribution of concrete to shear (V
c
) was
assessed according to NTCC-04 considering an effective
depth.
(b) The transverse steel reinforcement was placed at the maxi-
mum spacing allowed by NTCC-04, s
max
= d/2. The separation
used for the construction of the specimens was obtained
from adjusting the theoretical s
max
= 205 mm to s
max
= 185 -
mm, in order to have stirrups almost equally spaced at the
haunched length L
h
= 933 mm (Fig. 3), and
(c) additional shear reinforcement at the vertex zone was
placed to account for the abrupt change of direction of the
bottom longitudinal reinforcement (Fig. 4). For this purpose,
a recommendation originally proposed by Park and Paulay
Fig. 1. Urban bridge with RC haunched beams in downtown Seattle.
Fig. 2. Urban bridges with RC haunched beams in Buenavista District, Mexico City
(photo courtesy of Luis A. UrbinaCalifornias).
h
L 25 = 93.3 93.3 93.3 25
25 L= 280
330
25
h
m
a
x
=

4
5
= 0, 3.07, 6.12,
9.13 and 12.10

5
a = 108.3
V V
10
(
4
5
,

4
0
,

3
5
,

3
0

a
n
d

2
5
)
[cm]
=

v
a
r
i
a
b
l
e

+ +
V - - V
m
i
n
h
Fig. 3. Geometry and loading for the test specimens.
28 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
[4] was adapted. According to Fig. 4, the additional vertical
(shear) force that can be transmitted at the vertex at yielding
is:
F
v
T
y
sina A
s
f
y
sina 1
where A
s
is the area of the longitudinal inclined steel reinforcement
and a is the angle of slope of haunch from horizontal. Therefore, the
separation of the additional shear reinforcement at the vertex re-
gion (s
cd
) can be computed as:
s
cd

A
v
f
y
d
min
F
v
2
where d
min
is the effective depth at the prismatic section of the
haunched beam and A
v
is the area of the transverse shear reinforce-
ment. The theoretical separation s
cd
was computed according to Eqs.
(1) and (2) and then adjusted to practical values in order to ease the
construction of the specimens. If the computed or adjusted s
cd
was
smaller than the maximum separation for the transverse steel rein-
forcement (s
max
= 185 mm), the transverse reinforcement at the ver-
tex zone was spaced at s
cd
; otherwise, the transverse reinforcement
at the vertex zone was spaced at s
max
.
The shear capacity at the haunched sections was later checked
with semi-empirical equations previously proposed by the authors
[14], as described later in this paper in the section comparison
with nominal shear strength.
The cryptogram for the identication of the prototypes is TAS-
Cai-Rj-c, where i is an index that indicates the considered
haunched angle: i = 0 = 0, i = 1 = 3.07, i = 2 = 6.12, i = 3 = 9.13,
and i = 4 = 12.10; j is an index that indicates the shear reinforced:
j = 0 indicates the absence of shear reinforced whereas j = 1
indicates the use of a minimum shear reinforced with the charac-
teristic already referred.
Following the described general design procedure, the provided
longitudinal reinforcement for all beams consisted of 3#8 bars at
the top and 4#8 bars at the bottom (#8 bars = bars one inch in
diameter). For all beams with shear reinforcement, the transverse
shear reinforcement consisted of seven closed stirrups (two legs)
made with #2.5 bars (#2.5 bars = bars 5/16 inch in diameter) along
the haunched length (7S#2.5@ 185 mm) and 2S#2.5@ 185 mm at
the prismatic length. The reinforcement at the vertex zone con-
sisted of 3S#2.5@ 185 mm for beams up to 3.07 (TASCa0-R1-c
and TASCa1-R1-c), 3S#2.5@ 140 mm for beam TASCa2-R1-c,
3S#2.5@ 75 mm for beam TASCa3-R1-c and 3S#2.5@ 45 mm for
beam TASCa4-R1-c. Typical arrangements are shown in Figs. 57.
3. Instrumentation and test displacement history
In order to assess the contribution of the steel reinforcement,
beams were internally instrumented with strain gages to measure
tensional and compressional strains in the longitudinal steel rein-
forcement along the haunched length, as well as to measure strains
of some stirrups in the same zone, as schematically depicted in
Figs. 8 and 9.
Beams were simply supported and tested under concentrated
cyclic loads (V) that were applied 100 mm from the vertex formed
by the intersection of tapered sections with the prismatic section
towards the centerline, as depicted in Fig. 3 and 10a. Loads and
reactions were applied through 25.4 100 220 mm steel plates
to avoid local bearing crushing failures. The load was measured
with load cells at each point of loading.
Table 1
Measured experimental properties for the steel reinforcement.
Bar number f
y
kg/cm
2
(MPa) e
y
f
sh
kg/cm
2
(MPa) e
sh
f
u
kg/cm
2
(MPa) e
u
8 4348 (426.5) 0.00237 4348 (426.5) 0.0086 7707 (756.1)
2.5 4592 (450.5) 0.00235 4592 (450.5) 0.0074 7436 (729.5)
Table 2
Measured experimental shear forces and characteristic displacements from cyclic tests.
Beam ID a () h
+
() f
0
c
kg=cm
2
MPa V
cr
t (kN) V
u
t (kN) V
collapse
t (kN) V
nHB
t (kN) V
nHB
/V
u
d
cr
(mm) d
u
(mm) d
collapse
(mm)
TASCa0-R0-c 0 51 454 (44.5) 8.78 (86.1) 12.61 (123.7) 12.06 (118.3) 9.94 (97.5) 0.788 6.06 12.10 18.10
TASCa1-R0-c 3.07 43 433 (42.5) 4.41 (43.3) 4.41 (43.3) 4.41 (43.3) 8.58 (84.2) 1.946 6.12 9.48 19.00
TASCa2-R0-c 6.12 41 354 (34.7) 5.97 (58.6) 6.08 (59.6) 4.21 (41.3) 6.69 (65.6) 1.100 6.14 8.88 24.30
TASCa3-R0-c 9.13 35 395 (38.7) 3.83 (37.6) 3.85 (37.8) 4.37 (42.9) 5.98 (58.7) 1.553 6.10 9.22 30.22
TASCa4-R0-c 12.10 33 361 (35.4) 1.61 (15.8) 2.76 (27.1) 3.41 (33.5) 4.67 (45.8) 1.692 3.08 18.14 60.46
TASCa0-R1-c 0 38 227 (22.3) 4.56 (44.7) 24.89 (244.2) 12.92 (126.1) 17.88 (175.4) 0.718 4.14 20.42 33.60
TASCa1-R1-c 3.07 40 245 (24.0) 8.18 (80.2) 20.75 (203.6) 11.38 (111.6) 16.05 (157.5) 0.733 8.04 24.10 36.46
TASCa2-R1-c 6.12 33
o
217 (21.3) 6.16 (60.4) 13.23 (129.8) 7.55 (74.1) 13.73 (134.7) 1.038 8.00 20.02 32.40
TASCa3-R1-c 9.13 36 284 (27.9) 2.92 (28.6) 13.70 (134.4) 9.77 (95.8) 12.43 (121.9) 0.907 4.06 29.28 40.56
TASCa4-R1-c 12.10 29 245 (24.0) 1.52 (14.9) 7.88 (77.3) 5.12 (50.2) 9.94 (97.5) 1.261 4.08 33.92 57.88
Fig. 4. Additional shear reinforcement at the vertex zone due to the abrupt change of direction of the bottom longitudinal reinforcement.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 29
Author's personal copy
External instrumentation for cyclic loading was designed to
measure vertical deections at midspan (two micrometers, one
to measure deections of beams with respect to the reaction beam
and one to measure the deection of the reaction beam with re-
spect to the ground oor, to correct beam deections) and the ap-
plied loads with four load cells (Fig. 10a).
Cyclic tests were displacement-controlled in terms of the mea-
sured displacement at midspan d. According to previous results of
monotonic tests [14], in order to capture rst shear cracking and
failure states, midspan displacement increments of 3 mm were
set in the displacement history in beams without shear reinforce-
ment, whereas in beams with stirrups, displacement increments
of 4 mm were used. Positive loads (gravity direction) induce a
positive moment and vice versa. Two cycles at the same displace-
ment were set in the displacement history, as schematically de-
picted in Fig. 10b. This was done in order to evaluate key
structural parameters such as stiffness and strength degradation,
energy dissipation, and equivalent viscous damping. Tests were
stopped when beams lost the ability of supporting more load,
due to excessive damage (structural instability).
4. Experimental results
4.1. Hysteretic response
Hysteretic curves obtained for the test specimens without and
with shear reinforcement are depicted in Figs. 11 and 12 respec-
tively. The deection (d) at midspan was corrected by a xed co-
lineal measurement to take into account the exibility of the
strong beam where the hydraulic jacks reacted (Fig. 10a). The shear
force (V) corresponds to the haunched end where the failure
occurred.
Because prototypes were subjected to the same deection his-
tory, which means that deection history is not a variable [19], it
is feasible to describe a general behavior for the RCHBs: increasing
the haunched angle diminishes the shear capacity and the stiffness
of the beams (directly related with the fact that when the
haunched angle increases, the volume of concrete diminishes
[14]), but increases the number of cycles that can sustain and their
deformation capacity (primarily related to the capacity of RCHB to
redistribute cracking along the haunched length). All elements
exhibited a pinching at the origin which is characteristic of
25 93.3
SECTION 1
93.3
SECTION 1
330
93.3
3 # 2.5 @ 4.5
4 # 8
3 # 8
SECTION 2
25
[cm]
8 # 2.5 @ 18.5
Fig. 6. Reinforcement for beam TASCa4-R1-c.
4 # 8
3 # 8
SECTION 2
# 2.5
SECTION 1
4 # 8
22
35 45
[cm]
25
30
22
40
45
3 # 8
VARIABLE
4
4
4
4
VARIABLE
30
35
40
45
25
STIRRUPS
STIRRUPS
# 2.5
Fig. 7. Typical cross sections.
Fig. 8. Typical internal instrumentation for beams TASCai-R0-c.
25 93.3
SECTION 1 SECTION 1
93.3
2 # 2.5 @ 110
4 # 8
330
3 # 8
93.3
SECTION 2
25
[cm]
Fig. 5. Reinforcement for beam TASCa1-R0-c.
30 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
elements failing in shear [20]. As the haunch angle a increases, the
pinching behavior is more pronounced (Figs. 11 and 12) due to the
sliding along shear cracks. In addition, it is evident a slight asym-
metry in the hysteresis due to: (1) the geometry and reinforcement
asymmetry of the prototypes with respect to a longitudinal axis
and (2) the self weight of the beams.
On the basis of experimental observations and in agreement
with the monotonic testing [14], three characteristic forces were
identied from the full hysteretic response: (1) the shear force that
caused the rst diagonal cracking (V
cr
), (2) the ultimate (maxi-
mum) shear force (V
u
) and (3) the shear that caused the collapse
of the beams (V
collapse
). All characteristic stages occurred under
(a)
-24
-20
-16
-12
-8
-4
0
4
8
12
16
20
24
0 1 2 3 4 5 6 7 8 9 10 11
(mm)
first cycles
second cycles
(b)
CYCLE NUMBER
Fig. 10. Experimental program: (a) testing setup (prototype TASCa4-R1-c) and (b) cyclic displacement history.
Fig. 11. Hysteretic curves for beams TASCai-R0-c (without shear reinforcement).
Fig. 9. Typical internal instrumentation for beams TASCai-R1-c.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 31
Author's personal copy
positive loads (Fig. 3). The characteristic shear forces already de-
scribed are summarized in Table 2. Finally, the displacements
associated to each characteristic stage are also summarized in
Table 2.
Fig. 12. Hysteretic curves for beams TASCai-R1-c (with shear reinforcement).
Fig. 13. Loaddisplacement envelope curves for beams TASCai-R0-c (without shear reinforcement).
32 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
4.2. Peak response envelopes
Loaddisplacement envelopes for peak responses obtained for
the rst and second cycles of deformation (distinguished with dif-
ferent lines) are depicted in Fig. 13 for the test specimens without
shear reinforcement and in Fig. 14 for the test specimens with
shear reinforcement, where plots are in different scale to improve
visualization. Nominal shear strength (V
nHB
) capacities assessed as
described in the following section are also depicted.
In these curves it can be conrmed that as the haunched angle
increases: (a) the deformation capacity increases and, (b) the shear
strength is reduced. As it is explained with more detail in next sec-
tion, it is evident in Fig. 13 that the expected shear capacity in
beams without shear reinforcement (R0-c beams) was overesti-
mated, whereas in Fig. 14 it is observed that the prediction of the
shear capacity was much better for haunched beams with mini-
mum shear reinforcement.
It is worth noting in Fig. 13 that beam TASCa1-R0-c (3) pre-
sented an anomalous brittle failure at a smaller shear force than
TASCa2-R0-c (6), as in the monotonic tests it was always observed
that the shear strength was reduced as the haunched angle a in-
creased (Table 3). First diagonal cracking occurred in most beams
at a displacement around 6 mm (Table 2), except for beam TAS-
Ca4-R0-c (12), where rst cracking occurred at 3 mm.
It can also be observed from the loaddisplacement envelopes
depicted in Figs. 13 and 14 that the envelopes for rst and second
cycles start to differ after the major crack was formed at or near the
failure state, although this difference is smaller for beams TASCa1-
R0-c (3), TASCa4-R0-c (12) and TASCa4-R1-c (12). In other
words, stiffness and strength degradation due to cyclic loading
Fig. 14. Loaddisplacement envelope curves for beams TASCai-R1-c (with shear reinforcement).
Table 3
Evaluation of cyclic degradation from cyclic and monotonic test results.
Beam type a () Cyclic test results Monotonic test results Evaluation of the cyclic
degradation
Vuc

f
0
c
p

kg
p
cm
Vcolc

f
0
c
p

kg
p
cm
f
0
c
kg=cm
2
MPa
V
u-m
t (kN) V
col-m
t (kN)
Vum

f
0
c
p

kg
p
cm
Vcolm

f
0
c
p

kg
p
cm
Vuc
Vum
Vcolc
Vcolm
Vuc =

f
0
c
p
Vum=

f
0
c
p
Vcolc =

f
0
c
p
Vcolm=

f
0
c
p
Without shear
reinforcement
0 591.8 566.0 334 (32.8) 7.5 (73.6) 7.5 (73.6) 410.4 410.4 1.68 1.61 1.44 1.38
3.07 211.9 211.9 321 (31.5) 6.75 (66.2) 8.75 (85.8) 376.7 488.4 0.65 0.50 0.56 0.43
6.12 323.1 223.8 295 (28.9) 6.0 (58.9) 6.5 (63.8) 349.3 378.4 1.01 0.65 0.93 0.59
9.13 193.7 219.9 236 (23.2) 3.75 (36.8) 8.0 (78.5) 244.1 520.8 1.03 0.55 0.79 0.42
12.10 145.3 179.5 281 (27.6) 3.0 (29.4) 4.0 (39.2) 179.0 238.6 0.92 0.85 0.81 0.75
With shear
reinforcement
0 1652.0 857.5 315 (30.9) 25.0 (245.3) 25.5 (250.2) 1408.6 1436.8 1.00 0.51 1.17 0.60
3.07 1325.7 727.0 269 (26.4) 20.0 (196.2) 21.0 (206.0) 1219.4 1280.4 1.04 0.54 1.09 0.57
6.12 898.1 512.5 292 (28.6) 17.0 (166.8) 17.0 (166.8) 994.8 994.8 0.78 0.44 0.90 0.52
9.13 812.9 579.7 288 (28.3) 12.0 (117.7) 14.0 (137.3) 707.1 825.0 1.14 0.70 1.15 0.70
12.10 503.4 327.1 212 (20.8) 8.0 (78.5) 8.0 (78.5) 549.4 549.4 0.99 0.64 0.92 0.60
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 33
Author's personal copy
have a negligible impact almost up to the ultimate load V
u
, as the
curves for the rst and second cycles are almost identical; how-
ever, strength and stiffness degradation become very important
after V
u
.
Loaddisplacement envelopes for peak responses obtained for
the rst cycles of deformation for the test specimens with shear
reinforcement (TASCai-R1-c) are compared in Fig. 15 with those
obtained for their counterpart twin beams previously tested under
monotonic loading (TASCai-R1) [14]. From the observation of this
gure one may conclude that for most beams, the developed shear
capacity (V
u
) for monotonic and cyclic (positive cycles) tests are
very similar, and that the differences in the global deformation
capacity are also small. In addition, measured experimental ulti-
mate and collapse shear forces for cyclic (V
u-c
and V
col-c
) and mono-
tonic (V
u-m
and V
col-m
) tests are compared in Table 3 by computing
a simple ratio between the strength for cyclic and monotonic loads
(
Vuc
Vum
and
V
colc
V
colm
). The previous observation is also supported with the
obtained
Vuc
Vum
ratios for test specimens with shear reinforcement, as
the only beam where a reduction of strength higher than 20% was
observed was the 6.12 haunched angle beam TASCa2-R1-c.
Excluding prismatic beams, the average ratio for the four studied
haunched angles was
Vuc
Vum
0:99. The importance of the degrading
effects of cyclic loads after major cracking occurred can be
observed in Fig. 15 and can be addressed with the obtained
V
colc
V
colm
ratios for test specimens with shear reinforcement. Much higher
strength degradation was observed for all specimens subjected to
cyclic loads. Excluding prismatic beams, the average ratio for the
four studied haunched angles was
V
colc
V
colm
0:58.
Perhaps a one to one comparison between cyclic and monotonic
tests might not be fair enough, as the concrete mechanic character-
istics (i.e., f
0
c
, Table 3) are different because: (a) concrete admix-
tures were not made at the same time, as specimens for
monotonic tests were casted several months before, (b) the age
of testing was different for each specimen and (c) environmental
variables (temperature, humidity, etc.) may affect the strength
and stiffness for the concrete. For these reasons, measured experi-
mental shear forces (at ultimate and at collapse) for cyclic and
monotonic tests were normalized with the square root of the
compressive strength of concrete (f
0
c
) at the age of testing for each
element, as also identied in Table 3. The reported
Vuc =

f
0
c
p
Vum=

f
0
c
p and
V
colc
=

f
0
c
p
V
colm
=

f
0
c
p ratios mostly conrmed previous observations for test
specimens with shear reinforcement: (a) the developed shear
capacity for monotonic and cyclic tests were very similar and (b)
much higher strength degradation was observed for specimens
subjected to cyclic loads with respect to those subjected to
monotonic loads. Excluding prismatic beams, the average ratios
for the four studied haunched angles were
Vuc =

f
0
c
p
Vum=

f
0
c
p 1:01 and
V
colc
=

f
0
c
p
V
colm
=

f
0
c
p 0:60.
4.3. Comparison with nominal shear strength
As explained in greater detail in Tena-Colunga et al. [14],
nominal shear strength capacities in the haunched length (V
nHB
)
have been traditionally assessed according to an empirical but
statics-based approach [3,4,7,912], in which the shear resistance
of a section is the sum of the contribution of the concrete, plus
the contribution of inclined and transverse steel reinforcements.
This methodology is the most widely used and accepted by
practicing engineers and that is the reason that still prevails in
most reinforced concrete regulations worldwide.
Nominal shear strength (V
nHB
) capacities were computed
according to what it is proposed in Tena-Colunga et al. [14] are also
reported in Table 2:
V
nHB
V
pc
V
isr
V
s
3
where V
pc
is the concrete shear strength of constant depth beam
having the same cross section as that considered for haunched
beam, V
isr
is the shear contribution of the inclined steel reinforce-
ment and V
s
is the contribution of the transverse steel reinforce-
ment. In the assessment of V
isr
, the positive sign should be used if
the moment increases in the same direction as the depth increases,
and the negative sign when the depth decreases while the moment
increases [7,914]; for the simply supported haunched beams un-
der study, a negative sign must be used. In Eq. (3), V
pc
and V
isr
terms
were obtained from the moment equilibrium equation at the sec-
tion of interest.
Individual shear contributions were assessed as follows:
V
pc
0:5

f
c
p
176q
w
V
n
d
cr
M
n

bd
cr
4
V
isr

M
cr
d
cr
tana 5
V
s

A
v
f
yv
d
cr
s tan45 a
6
d
cr
d
min
1 1:35tana 6
h
max
h
min
h
2
max
2l
h
h
max
!
r
" #
7
-28.5
-14.3
0.0
14.3
28.5
-280
-140
0
140
280
-60 -40 -20 0 20 40 60
V (t) V (kN)
(mm)
0
1
2
4
3
-28.5
-14.3
0.0
14.3
28.5
-280
-140
0
140
280
0 20 40 60
V (t) V
(kN)
(mm)
0
1
2
3
4
(a) Cyclic tests (TASCi-R1-c) (b) Monotonic tests (TASCi-R1)
Fig. 15. Comparison of loaddisplacement envelope curves for beams with shear reinforcement.
34 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
where V
n
and M
n
are the nominal shear force and bending moment
at the critical haunched section (absolute values), q
w
is the ratio of
longitudinal reinforcement, b is the width of the beam, d
cr
is the
effective equivalent depth at the critical haunched section, M
cr
is
the developed bending moment at the critical section (taking into
account premature bond-slip failures of the inclined reinforcement
Fig. 16. Final damage patterns for beams TASCai-R0-c (without shear reinforcement).
Fig. 17. Cracking patterns in beam TASCa0-R1-c at dened damage states.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 35
Author's personal copy
in over-reinforced beams), a is the angle of slope of haunch from
horizontal, h
max
and h
min
are the maximum and minimum depth
of the haunched beam respectively, L
h
is the haunched length, r is
the concrete cover for the longitudinal reinforcement, A
v
is the area
of the transverse shear reinforcement, f
yv
is the yield stress of the
shear reinforcement and s is the separation of the transverse shear
reinforcement.
It is worth noting that for consistency with prismatic sections,
V
pc
(Eq. (4)) was derived in terms of the equation proposed by
the ACI 318 code [2], whereas V
isr
(Eq. (5)) is the contribution of
the inclined steel reinforcement as already proposed in the litera-
ture [7,914]. The contribution of the transverse steel reinforce-
ment V
s
(Eq. (6)) was derived considering the angle of inclination
of the principal shear crack with respect to an horizontal axis (h),
which was proposed as h = 45-a based upon experimental evi-
dence, practical considerations and for consistency with the design
already established for prismatic sections [14]. The rst term of d
cr
(Eq. (7)) was obtained from a linear regression of experimental
data [14], whereas the limiting value provided in the second term
of Eq. (7) was derived analytically to insure that d
cr
would keep a
physical meaning, avoiding that it could be numerically greater
than the maximum effective depth [14].
Based upon previous experimental data [14], for simply sup-
ported beams without shear reinforcement, the assessment of
the developed moment at the critical section (M
cr
) could be done
as follows:
M
cr
0:15M
n
1 2:5tana 8
and for simply supported beams with shear reinforcement:
M
cr
0:5M
n
1 1:6tana 9
Then, shear capacities V
nHB
were computed considering the mea-
sured properties for the concrete and steel reinforcement. The mea-
sured compressive strength for the concrete (f
0
c
) is reported in Table
2. The measured yield stresses were f
y
= 426.5 MPa (f
y
= 4348 kg/
cm
2
) for the longitudinal reinforcement and f
y
= 450.5 MPa (f
y
=
4592 kg/cm
2
) for the shear reinforcement (Table 1).
From the reported values in Table 2 it is evident that the
expected shear capacity in beams without shear reinforcement
(R0-c beams) was overestimated. This can be explained by the fact
that the absence of stirrups favored a more pronounced concrete
degradation that was increased by load reversals [21]. As observed
in monotonic testing [14] for the haunched beams with shear rein-
forcement, the shear capacity prediction was good principally in
Fig. 18. Cracking patterns in beam TASCa1-R1-c at dened damage states.
36 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
haunched beams up to haunched angles of nine degrees for the
beams with minimum shear reinforcement. The average ratio for
all ve beams with transverse shear reinforcement was V
nHB
/
V
u
= 0.94 with and standard deviation of 0.218. However, if only
the four haunched beams are considered, the average ratio in-
creases to V
nHB
/V
u
= 0.995 and the standard deviation decreases
to 0.208. Beam TASCa2-R1-c (a = 6.12) was tested at early age
(32 days old) in comparison with the rest of the beams (200
250 days old, because the testing lab in the university was closed
as it was going to be upgraded). In addition, this element had the
lowest concrete strength (Table 2). These facts favored an excessive
exural cracking along the element which could explain the devel-
oped strength.
4.4. Deformation capacity
It can be observed from the loaddisplacement envelopes for
peak responses depicted in Figs. 13 and 14 and the data provided
in Table 2 that curves for beams with (R1-c) and without (R0-c)
shear reinforcement are similar until V
cr
, but after the rst crack-
ing, the strength and deformation of R0-c and R1-c specimens start
to differ because of the contribution of the steel shear reinforce-
ment. It is also evident from Fig. 13 that RCHB have some deforma-
tion capacity even without shear reinforcement, situation different
from prismatic beams, where once the diagonal cracking occurred,
prismatic beams lose all capacity to sustain the load through defor-
mation. This increment is primarily related to the capacity of RCHB
to redistribute cracking along the haunched length, as discussed in
more detail in the following section.
4.5. Cracking patterns
As mentioned earlier, haunched beams exhibited a greater
deformation and energy dissipation capacity. This increment is pri-
marily related to the capacity of RCHB to redistribute cracking
along the haunched length (Fig. 16b) compared to prismatic beams
(Fig. 16a), even for beams without shear reinforcement. The better
cracking distribution (i.e., Fig. 16b) allowed that the ultimate shear
strength V
u
did not suddenly happened after the appearance of the
rst important diagonal crack, as it happened for prismatic beams
in both monotonic and cyclic tests (Fig. 16a). The observed behav-
ior for RCHB is less brittle than for prismatic beams, as the shear
failure for RCHB is noticeably less sudden than the one presented
in prismatic beams. These observations are in full agreement with
the results obtained for the twin beams during monotonic tests
[14].
Typical cracking patterns for beams with shear reinforcement
are shown in Figs. 1721. For the RCHBs (Figs. 1821), the cracking
patterns at ultimate and collapse stages clearly show a full diago-
nal crack distribution along the shear span (haunched length). This
crack pattern supports the well identied arching mechanism in
monotonic tests [914] and analytical studies. This arching mech-
anism was pronounced by the presence, in each haunch, of inclined
compression struts in both loading directions. In agreement with
monotonic tests [14], these struts tend to form between the point
Fig. 19. Cracking patterns in beam TASCa2-R1-c at dened damage states.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 37
Author's personal copy
of application of the load and the midpoint of the haunched length.
As expected, deformation demands were higher in the struts of
RCHB with shear reinforcement, where the force developed by
the strut considerably deformed the initial geometry of the beam
(Figs. 18c, 19c, 20c and 21c). Therefore, it can be concluded that
the greatest deformation capacity observed in RCHBs in compari-
son with the prismatic elements is associated to the ability of
RCHBs to redistribute cracks. As a general rule, increasing the
haunched angle increases the allowed damage (Figs. 1721) and
increases the pinching (Fig. 12) due to the sliding along the cracks.
Moreover, the characteristic brittle and sudden shear failure of
prismatic elements (Fig. 17) is reduced in RCHBs due to this behav-
ior (Figs. 1821).
As summary, the cracking patterns identied in the tested
RCHBs under cyclic loading conrmed the arch mechanism previ-
ously reported for RCHBs with shear reinforcement [10,14] and
without shear reinforcement [13,14]. This arch mechanism was
pronounced by the presence, in each haunch, of inclined compres-
sion struts. The struts tend to form between the point of applica-
tion of the load and the midpoint of the haunched length. As
expected, deformation demands were higher in the struts of RCHBs
with shear reinforcement, where the force developed by the strut
considerably deformed the initial geometry of the beam. For all
RCHBs, the shear cracking was distributed, and more than one
important diagonal crack always appeared with or without shear
reinforcement.
Finally, it was conrmed that as observed in monotonic tests
[14], there is a trend where the angle of inclination of the principal
shear crack with respect to a horizontal axis (h
+
, Table 2) seem to
decrease as the haunch angle a increases.
4.6. Longitudinal reinforcement
In agreement with previous monotonic tests [14], readings of
the strain gages in the inclined longitudinal steel reinforcement
in tension showed that the deformations associated to the failure
of the specimens without shear reinforcement were much smaller
than the yield strain of the steel reinforcement obtained from
coupon tests (about 12%). These values were obtained usually
between V
u
and V
collapse
(Table 2) and a displacement around
1213 mm, before readings were lost when the opening of a major
crack damaged the strain gage. The longitudinal steel reinforce-
ment at the top had similar readings than the ones obtained for
the inclined reinforcement; however, as they were not damaged
because of the presence of a major crack, they kept giving readings
up to displacements near d
collapse
. For example, for beam TASCa4-
R0-c, at d = 55.6 mm, the peak strain reading was about 56% the
yield strain of the steel reinforcement.
The applied shear force (V) was then related to the measured
strain of the top and bottom longitudinal steel reinforcement for
all specimens as described in detail in Grande-Vega [22]. As ex-
pected, higher strains developed at the steel reinforcement from
beam ends towards the points of application of the concentrated
loads, as the bending moment increases in this direction. Yielding
of the steel reinforcement (e P0.00213) was measured at strain
gages located at the vertex and at the tapered section nearby the
vertex and at midspan. It was also evident that the development
of crack patterns affected the readings of the strain gages, as one
can relate the developed crack patterns (Fig. 20) with the curves
obtained from the measured strains (not shown, Grande Vega
[22]).
4.7. Shear reinforcement
The contribution of the shear reinforcement of R1-c specimens
was quantied with the measured deformations in the stirrups as
reported in detail elsewhere [22]. For illustration purposes, the ap-
plied shear force (V) versus measured strains in the stirrups curves
for the haunched element TASCa3-R1-c is depicted in Fig. 22. It is
worth noting that stirrup E03-I had wrong readings, perhaps be-
Fig. 20. Cracking patterns in beam TASCa3-R1-c at dened damage states.
38 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
cause it was deciently glued it to the steel and then got lose dur-
ing the testing. It can be observed that all stirrups were working
primarily in tension, although most of them did not yield
(e < 0.00235) but stirrups nearby the vertex (or point of application
of the loads). In fact, higher strains are developed at the stirrups
from the beam ends towards the vertex. The most demanded stir-
rups were those that were located between the vertex and the
midsection of the haunched length (that is, stirrups E04-I, E05-I,
E06-I, E07-I at the left hand side and stirrups E04-D, E05-D, E06-
D, and E07-D at the right hand side). This is consistent with the
concentration of the damage in that zone because of the compres-
sion concrete struts and the location of the principal shear crack
(Fig. 20). It is worth noting that, as expected, stirrups at the sup-
ports (E01-I and E01-D) were not very demanded. Based upon
the analysis of the cyclic testing sequences, it was found that mea-
sured strains in the shear reinforcement (Fig. 22) were also affected
by the development (sequence) of the crack patterns (Fig. 20).
4.8. Stiffness degradation
Peak to peak secant stiffness (K) was computed for each com-
plete cycle (Fig. 23) and normalized with respect to the initial elas-
tic stiffness (K
0
), which was computed from the rst cycle readings
at small displacement amplitude. Elastic and normalized stiffness-
es are summarized in Table 4 whereas the evolution of the stiffness
degradation is depicted in Fig. 24. The effective secant stiffness for
the rst cycles of deformation was in general similar but slightly
higher than for the second cycles of response for all beams [19]
(not shown), and since the displacement amplitude of the cycles
was xed, this very small stiffness degradation must arise from
strength degradation. As expected, stiffness and strength degrada-
tion were more pronounced after the ultimate damage state (V
u
and d
u
, Table 2) was reached.
Indeed, as previously mentioned, stiffness and strength degra-
dation due to cyclic loading have a negligible impact almost up
to the ultimate load V
u
, as the curves for the rst and second cycles
are almost identical, but strength and stiffness degradation become
very important after V
u
(Figs. 13 and 14). In fact, before reaching V
u
,
a lack of stiffness and strength degradation due to previous cycles
of smaller amplitude and due to cycles of the same amplitude was
observed. However, near to and after V
u
, important stiffness and
strength degradation were observed due to: (a) previous cycles
of smaller amplitude and (b) cycles of the same amplitude.
From the results reported in Table 4 and the curves depicted in
Fig. 24 it is evident, and obvious, that the absence of stirrups in-
creases the stiffness degradation rate. It is specied in most rein-
forced concrete building codes [2,3], the Mexican code [1]
included, that a cracked beam shall to remain at 50% of their initial
Fig. 21. Cracking patterns in beam TASCa4-R1-c at dened damage states.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 39
Author's personal copy
stiffness properties at the moment to develop their exural capac-
ity. As can be observed from Table 4, RCHBs would be able to fulll
this requirement, because even under a brittle failure (shear
failure) the beams kept 60% of their original stiffness at the ulti-
mate stage.
4.9. Energy dissipation
Even though the haunch diminishes the concrete volume re-
spect to a prismatic element, the experimental evidence shown
that RCHBs dissipate, at least, energy equal to the prismatic ele-
ments (Table 5). To further highlight this fact, the accumulated en-
ergy density U
o
[23] was assessed, this is, the accumulated
E
-
0
8
-

I
E
-
0
7
-

I
E
-
0
6
-

I
E
-
0
5
-

I
E
-
0
4
-

I
E
-
0
3
-

I
E
-
0
2
-

I
E
-
0
1
-

I
E
-
0
1
-

D
E
-
0
2
-

D
E
-
0
3
-

D
E
-
0
4
-

D
E
-
0
5
-

D
E
-
0
6
-

D
E
-
0
7
-

D
E
-
0
8
-

D
Fig. 22. Measured strain versus shear curves for the strain gages of the transverse reinforcement of haunched beam TASCa3-R1-c.
Fig. 23. Peak-to-peak secant stiffness K.
Table 4
Peak to peak stiffness.
Beam ID K
o
(ton/mm) K
cr
/K
o
K
u
/K
o
K
collapse
/K
o
Beam ID K
o
(ton/mm) K
cr
/K
o
K
u
/K
o
K
collapse
/K
o
TASCa0-R0-c 1.46 0.92 0.60 0.42 TASCa0-R1-c 1.22 1.0 0.86 0.27
TASCa1-R0-c 0.72 0.85 0.61 0.31 TASCa1-R1-c 1.01 0.97 0.77 0.29
TASCa2-R0-c 0.95 0.98 0.58 0.20 TASCa2-R1-c 0.73 0.90 0.75 0.31
TASCa3-R0-c 0.64 0.92 0.55 0.25 TASCa3-R1-c 0.63 1.0 0.66 0.37
TASCa4-R0-c 0.57 0.98 0.28 0.06 TASCa4-R1-c 0.35 1.0 0.67 0.32
40 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742
Author's personal copy
hysteretic energy at failure (E
H
) was normalized with respect to the
half-volume of each beam (V
beam
/2), to point out the haunched ef-
fect (geometric effect). Therefore, U
o
was calculated as:
U
0

R
ncycles
0
@E
H
V
beam
2
10
The resulting accumulated energy densities are depicted in Fig. 25. It
can be conrmed that the accumulated energy density tend to in-
crease as the haunch angle a increases, as a direct consequence that
the number of cycles also tend to increase as the haunch angle a
increases.
5. Concluding remarks
This paper presented the results and interpretations about the
testing of prototype reinforced concrete haunched beams (RCHBs)
with and without shear reinforcement designed to develop a shear
failure under cyclic loading. The haunched length at both beam
ends of the considered RCHB was one-third the effective span of
the beam, whereas the haunched angle (a) varied from 3 to 12.
It was conrmed from cyclic tests what it was observed in
monotonic tests [14]: RCHB develop an arch mechanism which
allows the damage to be distributed in terms of several ssures
along the haunched length before the main diagonal cracks
develops, then causing smoother cracking patterns and a failure
mechanism that is less fragile than the typical sudden shear failure
observed in prismatic beams. The presence of a haunch modies
important structural properties such as ultimate strength and
deformation capacity, stiffness and energy dissipation. RCHB have
indeed more deformation capacity in shear than prismatic beams
and dissipate more energy or at least energy equal to those dissi-
pated by prismatic beams. On the other hand, the stiffness and
strength of RCHB is smaller with respect to those of prismatic
beams of the same depth.
From strain gages measurements it was conrmed that for sim-
ply-supported haunched beams, the presence of stirrups contrib-
utes indirectly to develop a higher negative vertical force by the
inclined longitudinal steel reinforcement, then, lessen in some ex-
Fig. 24. Peak to peak stiffness (computed from the rst cycles at a given displacement amplitude): (a) beams without shear reinforcement and (b) beams with shear
reinforcement.
Table 5
Accumulated hysteretic energy.
Beam ID Half volume (m
3
) Cracking Ultimate Collapse
E
H
t-mm (kN-mm) Cycle number E
H
t-mm (kN-mm) Cycle number E
H
t-mm (kN-mm) Cycle number
TASCa0-R0-c 0.144 26.53 (260.3) 3 144.55 (1418.0) 7 455.59 (4469.3) 11.5
TASCa1-R0-c
a
0.133 8.99 (88.2) 3 36.78 (360.8) 5 189.15 (1855.6) 11
TASCa2-R0-c 0.123 8.66 (85.0) 3 55.37 (543.2) 5 413.88 (4060.2) 15
TASCa3-R0-c 0.113 4.94 (48.5) 3 30.69 (301.1) 5 564.20 (5534.8) 20
TASCa4-R0-c 0.102 0.39 (3.8) 1 116.21 (1140.0) 11 2291.85 (22483.0) 36.5
TASCa0-R1-c 0.144 7.88 (77.3) 1 338.87 (3324.3) 9 1345.47 (13199.1) 15
TASCa1-R1-c 0.133 33.74 (331.0) 3 499.63 (4901.4) 11 1612.12 (15814.9) 17
TASCa2-R1-c
a
0.123 20.64 (202.5) 3 198.94 (1951.6) 9 904.19 (8870.1) 16
TASCa3-R1-c 0.113 4.53 (44.4) 1 491.53 (4821.9) 13 1418.89 (13919.3) 20
TASCa4-R1-c 0.102 2.34 (23.0) 1 425.13 (4170.5) 15 1739.62 (17062.7) 28
a
These elements had a premature failure.
Fig. 25. Accumulated energy density: (a) beams without shear reinforcement and (b) beams with shear reinforcement.
H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742 41
Author's personal copy
tent the positive effect of the same stirrups to resist higher shear
forces.
It was also found from cyclic tests that stiffness and strength
degradation have a negligible impact almost up to the ultimate
load V
u
, but they become very important after V
u
. In fact, before
reaching V
u
, a lack of stiffness and strength degradation due to pre-
vious cycles of smaller amplitude and cycles of the same amplitude
was observed. However, near to and after V
u
, important stiffness
and strength degradation were observed due to: (a) previous cycles
of smaller amplitude and (b) cycles of the same amplitude.
The results obtained from the cyclic testing also conrmed the
usefulness of an empirical equation previously proposed [14] to as-
sess the shear strength of reinforced concrete haunched beams,
taking into account parameters such as the haunch angle, the con-
crete compressive strength, the shear reinforcement and the con-
tribution of the inclined longitudinal reinforcement. Therefore,
until more evidence would be available, it is advisable to estimate
the shear strength of simply supported RCHBs with that equation
for haunched beams with a haunch angle a up to nine degrees.
This study constitutes a step forward towards the understand-
ing of the behavior of RCHB under shear which may allow develop-
ing, in the near future, coherent guidelines to prevent brittle shear
failures. However, additional analytical and experimental research
is still needed to try to achieve this goal, mainly: (a) development
of truss (strut-and-tie) and compression eld analytical models, (b)
additional testing must be conducted in RCHB where the maxi-
mum shear force and bending moment occurs at the same section
(continuity condition) for cyclic loading, as this condition is closer
to the one that RCHB in framed buildings must face under lateral
loading.
Steps are currently being taken in this direction, as they are in-
deed needed before starting experimental programs devoted to in-
sure the ductile behavior of RCHB designed to develop exural
failure under cyclic loading.
Acknowledgments
Financial support of Conacyt, Secretara General de Obras del
Gobierno del Distrito Federal and Universidad Autnoma Metro-
politana are gratefully acknowledged. The construction of test
specimens was possible because of the enthusiastic collaboration
of several students: Jos Ramn guila, Jos Manuel Alonso, Misael
Bahena, Marco Antonio Castillo, Csar Carpio, Arturo de la Cruz,
Eleuterio Flores, Antonio Gascn, Eder Gudio, Efran Joaqun Die-
go, Alejandro Herrera, Gerardo Ibarra, Mara Fernanda Juan, Sergio
Lpez, Daniel Miranda, Miguel ngel Mendoza, Elas Josu Morn,
Roberto Moreno, Pablo Pineda, Carlos Ramrez, Rosaura Ramrez,
Marco Antonio Rico, Martn Santos, Alejandra Salcedo, Salvador
Vargas and Francisco Javier Vidal. MSc. Artemio Jurez-Angeles
Vctor H. Salinas-Vallejo and technicians Leopoldo Quiroz, Juan
Mateos, Jos Luis Caballero and Rubn Barreda assisted us in the
prototype testing.
References
[1] NTCC-2004. Normas Tcnicas Complementarias para Diseo y Construccin de
Estructuras de Concreto Gaceta Ocial del Distrito Federal. Mxico, October;
2004 [in Spanish].
[2] ACI-318-11. Building code requirements for structural concrete (ACI-318-11)
and commentary (ACI 318R-11). American Concrete Institute, Farmington
Hills, Michigan, USA; 2011.
[3] DIN 1045-1. Concrete, reinforcement and prestressed concrete structures, part
1: design. Deutsches Institut fr Normung e.V.; 2001.
[4] Park R, Paulay T. Estructuras de concreto reforzado, ninth printing of the rst
edition. Limusa, Mxico; 1997. p. 27984, 6912 [in Spanish].
[5] MacGregor JG. Reinforced concrete: mechanics and design. 3rd
ed. USA: Prentice Hall; 1997.
[6] Muttoni A, Shwartz J, Thrlimann B. Design of concrete structures with stress
elds. Germany: Birkhuser; 1997.
[7] Nilson AH. Diseo de estructuras de concreto. 12th ed. Santaf de Bogot,
Colombia: Mc Graw Hill; 1999. p. 1323 (in Spanish).
[8] Nilsen MP. Limit analysis and concrete plasticity. 2nd ed. Boca Raton,
Florida: CRC Press; 1999. p. 435.
[9] Mrsch E. Teora y prctica del hormign armado. Tomo II, rst edition Gili,
Argentina; 1952 [translated to Spanish from German].
[10] Debaiky SY, El-Niema EI. Behavior and strength of reinforced concrete
haunched beams in shear. ACI J 1982;79(3):18494.
[11] Stefanou GD. Shear resistance for reinforced concrete beams with non-
prismatic sections. Eng Fract Mech 1983;18(3):64367.
[12] El-Niema EI. Investigation of concrete haunched beams under shear. ASCE
Struct J 1988;114(4):91730.
[13] MacLeod IA, Houmsi A. Shear strength of haunched beams without shear
reinforcement. ACI Struct J 1994;91(1):7989.
[14] Tena-Colunga A, Archundia-Aranda HI, Gonzlez-Cuevas OM. Behavior of
reinforced concrete haunched beams subjected to static shear loading. Eng
Struct 2008;30(2):47892. http://dx.doi.org/10.1016/j.engstruct.2007.04.017.
[15] Marti P. Basic tools of reinforced concrete beam design. ACI J
1985;82(1):4656.
[16] Dilger WH, Langorh P. Shear design of haunched concrete box girders of the
Confederation Bridge. Can J Civil Eng 1997;24(6):898907.
[17] Muttoni A. Exercice 6: Dimensionnement dune poutre hauteur variable.
Structures en bton I, Ecole Polytechnique Fdrale de Lausanne, ENAC, IS-
BETON, Switzerland; 2008 [in French].
[18] El-Mezaini N, Balkaya C, itipitioglu E. Analysis of frames with nonprismatic
members. ASCE J Struct Eng 1991;117(6):157392.
[19] Hwang T, Scribner CF. R/C members cyclic response during various loadings.
ASCE J Struct Eng 1984;110(3):47789.
[20] Brown RH, Jirsa JO. Reinforced concrete beams under load reversals. ACI J
1971;68(5):38090.
[21] Gosain NK, Brown RH, Jirsa JO. Shear requirements for load reversals on RC
members. ASCE J Struct Eng 1977;103(7):146176.
[22] Grande-Vega A. Mecanismos de resistencia y deformacin a cortante de trabes
acarteladas de concreto reforzado sujetas a cargas cclicas. MSc. Thesis,
Posgrado en Ingeniera Estructural, Divisin de Ciencias Bsicas e Ingeniera.
Universidad Autnoma Metropolitana Azcapotzalco, July 2009 [in Spanish].
[23] Popov EP. Engineering mechanics of solids. 2nd ed. USA: Prentice Hall; 1999.
42 H.I. Archundia-Aranda et al. / Engineering Structures 49 (2013) 2742