ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 96-S43

High-Strength Concrete Short Beams Subjected

to Cyclic Shear
by Yan Xiao, Asadollah Esmaeily-Ghasemabadi, and Hui Wu

Six high-strength concrete model short beams were tested under exhibited increased capacity and improved hysteretic perfor-
cyclic shear in a double curvature condition. Experimental mance when compared with the NSC beams.
parameters included reinforcement configurations, clear span Short or deep beams can be useful in tubular frame struc-
length-to-section depth ratios ranging from 3.0 to 4.0 (shear-span-
to-depth ratios of 1.5 to 2.0), and flexural reinforcement ratios
tures, coupled wall systems, foundations, etc. The benefits of
ranging from 2.3 to 4.5 percent. The stirrup design was based on using HSC in short reinforced concrete beams include the
the seismic design provisions of the ACI 318-95 Code. All the increased shear strength, increased bond strength, etc. Most
model beams developed flexural yielding prior to sliding shear recently, Tan et al.9 tested 19 reinforced concrete deep
failure corresponding to a displacement ductility factor of 3.5 to beams with compressive strengths ranging from 41 to 59
6.2. The flexural strength calculated based on the ACI 318-95 MPa (6000 to 8600 psi) under two-point top loading condi-
Code approach was developed or exceeded. Compared with beams tion. They confirmed that the ACI 318-95 shear design equa-
with conventional reinforcement configuration, the model beams tions for deep beams are applicable to HSC deep beams, but
with vertically distributed flexural reinforcement developed tend to be rather conservative for deeper beams with a shear-
significantly improved hysteretic performance and ductility.
span ratio less than 2.0.
The seismic design provisions of the ACI 318-95 Code limit
Keywords: ductility; flexural strength; high-strength concretes; shear
strength. the length-to-depth ratio of frame elements to not less than
4.0.8 Research literature shows that the existing studies on
seismic performance of short beams are focused on NSC short
INTRODUCTION
beams in coupled shearwall structures. Paulay studied the
The behavior of high-strength concrete (HSC) beams with
seismic behavior and failure modes of NSC short beams in
concrete strength higher than 41.4 MPa (fc′ > 4 ksi) has been
coupled shearwall structures10,11 and introduced the design
studied by many researchers.1-2 However, research into the
concept of using diagonal reinforcement in coupling beams.12
seismic behavior of HSC beams is still limited. Shin, Ghosh,
and Moreno3 tested a large number of small-scale HSC Recently, Theodosios et al. studied the behavior and
model beams subjected to monotonic or cyclic flexural ductility of NSC short coupling beams with various rein-
loading. They found that a deflection ductility factor of 4.0 forcement configurations including diagonal reinforcement
could be achieved by the model beams under cyclic loading. and vertically distributed bars.13 Paulay and Priestley14
Fang et al.4 investigated cyclic behavior of HSC beams with pointed out that beams with vertically distributed flexural
lower amounts of flexural reinforcement. In studies carried reinforcement may offer many advantages, such as: 1) easier
out by Shin et al. and Fang et al., the shear demand on the access in the top of the beam for placing and vibrating the
model beams was relatively low. Sugano et al.5 and Saka- concrete during construction; 2) better distribution and early
guchi et al.6 studied HSC beams under cyclic shear in double closing of flexural cracks on moment reversal compared
curvature condition. Excellent performance was reported for with conventional sections; 3) reduced tendency for exces-
the HSC beams reinforced with high-strength transverse sive sliding shear deformations in the plastic hinge region; 4)
reinforcement (fy > 800 MPa = 116 ksi). However, the high- increased depth of concrete compression zone in beam plastic
strength reinforcement is not commonly available elsewhere, hinges, thereby improving shear resistance; 5) smaller flexural
and whether cost-effective designs of HSC beams can be overstrength than conventionally reinforced sections under
achieved using current codes as well as available techniques larger curvature ductility demands; etc. Whether these advan-
and materials still needs to be investigated. tages, as evidenced in Wong et al.’s study15,16 and Theodosios
In an earlier phase of the research program described in et al.’s study13 on NSC beams, can be fully developed in HSC
this paper, Xiao and Ma7 studied the seismic performance of short beams is investigated in this study.
HSC beams with relatively longer spans. Two approximately Note that SI units are used as the main units throughout the
1:2 scale HSC and two counterpart normal strength concrete paper, with the corresponding English units provided in
(NSC) model beams with a length-to-depth ratio of 6.0 (or a parentheses. The sizes and grades of reinforcing bars are
shear-span ratio of 3.0) were tested under cyclic shear in described based on so-called soft conversions.17
double curvature. The beams were heavily reinforced in the
longitudinal direction to study beam performance under high
seismic shear demand. Transverse reinforcement was ACI Structural Journal, V. 96, No. 3, May-June 1999.
designed following the seismic design provisions of the ACI Received August 19, 1997, and reviewed under Institute publication policies. Copy-
right © 1999, American Concrete Institute. All rights reserved, including the making of copies
318-95 Code.8 Both normal and high-strength concrete unless permission is obtained from the copyright proprietors. Pertinent discussion including
author’s closure, if any, will be published in the March-April 2000 ACI Structural Journal if
beams developed ductile flexural responses. The HSC beams the discussion is received by November 1, 1999.

392 ACI Structural Journal/May-June 1999

 Table 1—Test matrix
ACI member Yan Xiao is an assistant professor of civil engineering at the University
of Southern California. He received his bachelor of engineering degree from Tianjin Length-depth Flexural
University, China, in 1982, and his MS and PhD degrees from Kyushu University, ratio (shear- Flexural steel ratio,† Transverse
Japan, in 1986 and 1989, respectively. He is a member of Joint ACI-ASCE Committee Specimen *
span ratio) reinforcement percent stirrups
441, Reinforced Concrete Columns. His research interests include earthquake-resistant
design of structures, structural concrete, steel, hybrid or composite systems, and HB4-6L-T100 Six No. 19 2.1 No. 10 at 100
structural materials. HB4-10L-T65 10 No. 19 3.4 No. 10 at 65
4.0 (2.0)
Asadollah Esmaeily-Ghasemabadi is a PhD candidate in civil engineering at the 12 No. 19
HB4-12L-T65 4.1 No. 10 at 65
University of Southern California. He received his BS and MS degrees from Tehran (Distributed)
University, Iran, in 1983 and 1986, respectively. His research interests include earth- HB3-6L-T100 Six No. 19 2.1 No. 10 at 100
quake-resistant design of reinforced concrete structures.
HB3-10L-T50 3.0 (1.5) 10 No. 19 3.4 No. 10 at 50
Hui Wu is a graduate research assistant in the Department of Civil and Environmental
HB3-12L-T50 12 No. 19
4.1 No. 10 at 50
Engineering at the University of Southern California. He received his BS in civil engi- (Distributed)
neering from Beijing Institute of Civil Engineering and Architecture in 1982, and his *Specimen designation: where H is high-strength concrete; B is beam; first number is
MS in structural engineering from Tsinghua University, China, in 1990. His research length-depth ratio; second number is number of bars used; L is No. 9 longitudinal
interests include high-strength concrete, prestressed concrete structures, and earth- bars; T is No. 10 transverse stirrups; final number is distance, in mm, of stirrups
quake-resistant design. spaced center-to-center.
†
Flexural steel ratio is total steel area divided by gross cross-sectional area of beam.
Note: 1 mm = 0.03937 in.

RESEARCH SIGNIFICANCE
The findings of this research contribute to the information
needed to develop future guidelines and code requirements deformed bars (i.e., ASTM No. 6; nominal diameter = 19.1
for seismic design of high-strength concrete structures. mm = 0.75 in.) and the other with 10 No. 19 bars. As the
Experimental evidence shows that adequate seismic counterpart specimen of the beam with 10 No. 19 bars in
behavior can be obtained for short HSC beams. Further each group, another beam was reinforced with 12 No. 19
improvement in seismic performance of heavily reinforced longitudinal bars that were distributed vertically with an
short HSC beams can be achieved using vertically distrib- equal number along the two sides of the beam section.
uted flexural reinforcement, as evidenced in this study.
A capacity design approach was followed in the determi-
nation of shear reinforcement. Using the equivalent concrete
EXPERIMENTAL PROGRAM
Model beam specimen design stress block and an ultimate strain of 0.003 recommended by
To study the behavior of HSC short beams under simu- the ACI 318-95 Code,8 the nominal moment strength Mn of
lated earthquake loading, six large-scale model specimens the HSC beam was first calculated. Then the probable flex-
were designed and constructed. Experimental parameters ural strength Mpr was taken as approximately 1.25Mn. The
include the clear span length-to-section depth ratio (shear- ultimate shear demands for the model beams were then
span ratio), flexural reinforcement contents, and configura- determined by the following Eq. (1)
tions. Table 1 summarizes the test matrix.
Based on the length-to-depth ratios, the six model beams 2M pr
can be divided into two groups. As shown in Fig. 1(a), all V u = -----------
- (1)
L
three model beams in the HB4 group had a length-to-depth
ratio of 4.0, which is approximately the minimum span limit
required by the seismic design provisions of the ACI 318-95 where Vu is required shear strength, and L is the clear span
Code.8 The HB3 group, as shown in Fig. 1(b), included three length of the beam.
shorter beams with a length-to-depth ratio of 3.0. All model Following the seismic shear design requirements of
beams had a section of 406 mm (16 in.) depth and 203 mm Chapter 21 in the ACI 318-95 Code,8 the concrete contribution
(8 in.) width. to shear strength was ignored, and the nominal shear strength
The ends of the model beams were framed into load stubs of a model beam Vn was thus determined based on the
for connecting to the loading arm and reaction floor beam following Eq. (2)
during testing. As shown in Fig. 1, the dimension of the load
stubs for the HB4 and HB3 groups was different, whereas the Av fy d
overall lengths of all six specimens were constructed the - ≥ Vu
V n = V s = -------------- (2)
s
same so that they could be tested using the same testing
configuration. The specimens can be considered to represent
1/2- to 2/3-scale models of prototype beams in actual cast in where Vn is the nominal shear strength; Vs is the shear
situ structures. The model beams were subjected to cyclic strength contribution by stirrups; Av is the cross-sectional
shear in double curvature with the points of inflection at their area of stirrups provided at spacing s; fy is the yield strength;
midspans. Thus, the length-to-depth ratios of 4.0 and 3.0 and d is the effective depth of the beam.
imply shear-span ratios of 2.0 and 1.5, respectively. Deformed No. 10 bars (i.e., ASTM No. 3; nominal diameter
Equal top and bottom reinforcements were provided = 9.5 mm = 3/8 in.) were used as stirrups. For beams with 10
throughout the lengths of the beams to simulate beam design or 12 No. 19 longitudinal bars, the spacing of No. 10 stirrups
dominated by severe earthquake loading. For the same was provided based on the above calculation. However, the
reason, relatively high contents of longitudinal steel were calculated spacing of No. 10 stirrups for model beams with
designed. In each group, two model beams were designed six No. 19 bars exceeded the maximum spacing requirement
with conventional flexural reinforcement with equal of one-fourth of the section effective depth, as specified in
numbers of longitudinal bars placed near the top and bottom the seismic provisions of the ACI 318-95 Code.8 Thus, the
of the section, as shown in Fig. 1(a) and (b). One of the stirrups in Specimen HB4-6L-T100 and HB3-6L-T100 were
conventional beams was reinforced with a total of six No. 19 spaced at 100 mm (4 in.) center-to-center. The ends of the

ACI Structural Journal/May-June 1999 393

 Fig. 1—Model beam details.

stirrups were anchored into the beam section using 135-deg cured in an air-dry condition after the removal of the forms.
hooks with 76-mm (3-in.) tail length. Each of the six specimens was tested within a period of 91 to
120 days after casting. Three 152 x 305-mm (6 x 12-in.)
Materials and construction cylinders were tested at 91 days and three more at 120 days.
The HSC was supplied by a local ready-mix concrete No meaningful difference between the strength at 91 days,
plant. Mix proportions for 1 m3 HSC were 187 kg water; 415 and that at 120 days was identified. An average strength of
kg cement; 148 kg Class F fly ash; 45 kg silica fume; 868 kg 69.5 MPa (10.1 ksi) was obtained.
coarse aggregate; and 710 kg fine aggregate. The water-to- The flexural reinforcing steel conformed to ASTM A 706
cementitious materials ratio was 30 percent. Superplasticizer with a yield strength of 510 MPa (74 ksi) obtained from labo-
was also used to improve workability and setting time. The ratory testing. Tensile tests of sample reinforcing bars indi-
average slump at casting was about 150 mm (about 6 in.). cated that the No. 10 bars possessed a yield strength of 469
The specimens were cured in the forms with their tops MPa (68 ksi). The reinforcing bars were manufactured by a
covered by wet burlap for a week after casting, and then were local reinforcing steel producer. Steel cages and the form-

394 ACI Structural Journal/May-June 1999

 Fig. 3—Loading program.

beam span, as shown in Fig. 2(a). The strains of selected

reinforcing bars were monitored by means of electrical resis-
tance gages mounted on the surfaces of the longitudinal and
transverse bars.

Loading program
The standard lateral loading program attempted for all
tests is shown in Fig. 3. Prior to testing each beam, the shear
corresponding to the first yield of the longitudinal reinforce-
Fig. 2—Test setup: (a) specimen and loading frame; and
ment Vy and the shear corresponding to the ideal flexural
(b) loading condition.
strength Vif were estimated for the specimen based on the
following equations
work were fabricated by professional carpenters. For
construction convenience, all beams were cast horizontally 2M
with one side facing upward. V y = ----------y (3)
L
Test setup
The tests were carried out using the newly constructed 2M
V if = ----------if- (4)
large-scale testing facility at the University of Southern Cali- L
fornia, which consists of a self-reacting steel and concrete
hybrid reaction frame. As shown in Fig. 2(a), the beam spec- where Mif is the estimated ideal flexural strength based on
imen was tested in the 90-deg rotated position with one end actual material strength, the equivalent concrete stress block,
fixed to the reaction floor beam and the other end to an L- and an extreme fiber strain of 0.003 as recommended by the
shaped loading arm. A double-acting 1334 kN (300 kips) ACI 318 Code,8 and My is the moment corresponding to the
capacity actuator was connected to the loading arm. The first yield of the flexural reinforcement.
loading system displaces the specimen in double curvature The loading cycles were controlled by peak lateral force
bending, with the point of inflection occurring at the increments of 22.2 or 44.5 kN (5 or 10 kips), up to the lateral
midspan of the beam, as schematically shown in Fig. 2(b). force equal to the estimated first yield strength. One loading
To assure the consistency of the loading condition and to cycle for each peak lateral force level was applied during
prevent accidental overturning of the L-shaped frame, a initial loading. Subsequent loading was carried out under
specially designed pantograph system was connected displacement control, with three attempted cycles corre-
between the loading arm and the reaction floor beam to sponding to each of the peak displacement ductility factors
provide restraint against rotation of the loading arm while of μΔ = 1, 1.5, 2, 3, or larger. The displacement ductility
allowing free vertical and horizontal displacements. The factor μΔ was defined as the ratio of peak displacement in
extra moment produced by the self-weight of the loading any cycle to the yield displacement Δyi corresponding to the
arm was counterbalanced by a lever system with a small ideal flexural strength, found from
dummy weight.
Relative deflections between the two ends of the beam V
were measured using two linear potentiometers with a stroke Δ yi = -----if- Δ 1 (5)
of 254 mm (10 in.). To evaluate shear and flexural defor- Vy
mations, longitudinal, transverse, and diagonal deformations
were measured using linear potentiometers with stroke of 38 where Δ1 is the average of the measured displacements for
mm (1.5 in.) for two or three equal length portions along the push and pull directions recorded at the first yield force Vy.

ACI Structural Journal/May-June 1999 395

 Fig. 4—Crack patterns at peak ductility levels.

EXPERIMENTAL RESULTS AND DISCUSSION Compared with their conventionally designed counterpart
General observations beams, the model beams with vertically distributed flexural
Crack patterns at various loading stages for all the six reinforcing bars developed finer cracks that were distributed
specimens are summarized in Fig. 4. For all six short beams, more evenly throughout the beam length, particularly in the
flexural cracks were observed during initial loading cycles case of HB3-12L-T50, as shown in Fig. 4(f).
corresponding to lateral force, less than 50 percent of the
calculated ideal flexural strength. After this stage, indepen- Although extensive numbers of inclined shear cracks
dent shear cracks started to occur in most specimens. The developed in all six specimens, no dominant shear crack
angle of inclination of the shear cracks to the beam axes causing rupture of shear reinforcement was observed. All
appeared to be approximately 45 deg. For beams with model beams developed the calculated flexural strengths,
length-to-depth ratio of 3.0 (or shear-span ratio of 1.5), the with yielding of flexural reinforcements verified through
shear cracks spread over the entire length of the beam span. strain gage readings. The beams lost their load-carrying

396 ACI Structural Journal/May-June 1999

Fig. 5—Hysteretic responses for model beams with six No. Fig. 6—Hysteretic responses for model beams with different
19 bars and different aspect ratios: (a) conventional reinforce- flexural reinforcement configurations (L/D = 4.0): (a)
ment (Model HB4-10L-T65); and (b) distributed reinforcement conventional reinforcement (Model HB3-10L-T50); and (b)
(Model HB4-12L-T65). distributed reinforcement (Model HB3-12L-T50).

capacities eventually due to sliding shear failure near their Table 2—Summary of test results
ends, upon loading cycles corresponding to peak displacement
ductility factor of 3.5 to 6.2. Compared with the beams rein- Vmax, Vif , kN Vmax/(bd f c ′ ), Δu, Δu/L,
forced with conventional flexural reinforcement [Fig. 4(b) and Specimen kN (Vmax/Vif ) MPa mm percent μΔ
(e)], the beams with distributed flexural reinforcement HB4-6L-T100 185 165 (1.12) 0.31 74 4.6 6.2
[Fig. 4(c) and (f)] sustained larger displacement and more HB4-10L-T65 271 256 (1.06) 0.45 59 3.6 4.0
loading cycles before sliding failure. HB4-12L-T65 296 289 (1.02) 0.49 82 5.0 5.7
HB3-6L-T100 242 222 (1.09) 0.40 44 3.6 6.0
Load-carrying capacities and ultimate HB3-10L-T50 354 342 (1.04) 0.59 39 3.2 3.5
deformations HB3-12L-T50 391 387 (1.01) 0.65 54 4.4 5.0
Table 2 summarizes the calculated shear corresponding to Note: Vmax is recorded maximum shear force; Vif is shear force corresponding to ideal
the ideal flexural strengths along with the recorded maximum flexural strength; Δu is ultimate lateral displacement achieved; L is clear span length;
shear forces, drift ratios, and ductility factors achieved by the and μΔ is ultimate lateral displacement ductility.

specimens. As shown in Table 2, the predicted ideal flexural

strength was developed by all six model beams. The ratios of
range from 3.5 to 6.2. The induced average shear stress in the
the experimental maximum shears to the calculated shears
beam section Vmax/bd ranges from 0.3 f c′ MPa (3.7 f c ′ psi)
corresponding to the ideal flexural strengths range from 1.01
to 1.12, indicating that the equivalent stress block approach of to 0.65 f c ′ MPa (7.8 f c ′ psi).
the ACI 318-95 Code8 is predictable but slightly conservative
for calculating flexural shear capacities of the model HSC Shear force-deflection hysteretic responses
short beams. The ultimate drift ratios (or rotational angles) The experimentally obtained shear force-versus-deflection
developed by the six HSC short beams range from 3.2 to 5.0 hysteretic responses for the six short beams are shown in
percent, whereas the ultimate displacement ductility factors Fig. 5 to 7. The primary horizontal axes in Fig. 5 to 7 indicate

ACI Structural Journal/May-June 1999 397

Fig. 7—Hysteretic responses for model beams with different
flexural reinforcement configurations (L/D = 3.0).
Fig. 8—Elongation responses for beams with different flexural
reinforcement configurations.
the drift ratios defined as the relative deflections between the
beam ends divided by the clear span length of the beam,
while the measured relative deflections are shown by the ratios of 4.0. However, the shortening in the span length does
secondary horizontal axes. The recorded lateral forces not necessarily imply a drastic decrease in the ultimate
measured in kN and kips are shown by the primary and the ductility factors, as indicated by comparing the counterpart
secondary vertical axes, respectively, in Fig. 5 to 7. specimens in the HB4 group and the HB3 group. By
Predicted shear forces Vif corresponding to the ideal flexural comparing the shapes of the hysteresis loops of the beams
strength, based on the equivalent rectangular compressive with different aspect ratios, it is clear that the pinching of the
stress block for concrete and an extreme fiber strain of 0.003, loops becomes more severe for shorter beams.
are shown by dashed lines. The ideal flexural strength was
calculated using actual material strengths as well as a unit Effects of longitudinal reinforcement
strength reduction factor. Ductility levels are indicated by configurations
vertical straight lines in Fig. 5 to 7. Hysteretic behavior for short HSC beams with conven-
tional design and the counterpart beams with vertically
Effects of length-to-depth ratio (shear-span ratio) distributed longitudinal reinforcement are shown in Fig. 6
The effects of the length-to-depth ratio or shear-span ratio and 7. The beams reinforced with vertically distributed
can be examined by comparing the hysteretic responses for longitudinal reinforcement were designed with a higher steel
the specimens in the HB4 group and the HB3 group, shown content to assure a higher shear demand than their counter-
in Fig. 5 to 7. The reduction in the span length-to-depth ratio part specimens with conventional design, although beams
from 4.0 (or shear-span ratio of 2.0) in the HB4 group to 3.0 with the same flexural reinforcement ratios but different
(or shear-span ratio of 1.5) in the HB3 group is reflected in a configurations would probably have approximately the same
proportional augmentation of load-carrying capacity. The flexural strengths, based on Wong et al.’s analysis.15 As
ultimate deflections and drift ratios achieved by the beams shown in Fig. 6 and 7, despite the increased lateral force
with length-to-depth ratio of 3.0 are about 15 to 25 percent carrying capacities (thus, shear demands), the HSC beams
less than those achieved by beams with length-to-depth with distributed longitudinal reinforcement developed

398 ACI Structural Journal/May-June 1999

significantly improved seismic behavior, characterized by Comparing the dashed with the solid curves in Fig. 8, it is
less pinching effects in the hysteresis loops and increased clear that beams with conventional flexural reinforcement
ultimate ductility levels, compared with their counterpart developed more significant elongation deformation than
conventional beams. Model Beam HB4-12L-T65 and HB3- beams with vertically distributed reinforcement, at the same
12L-T50 developed ultimate ductility factors of 5.5 and 5.0, lateral displacement levels prior to the sliding shear failure.
compared with 4.0 and 3.5 for their counterpart convention- The beam elongation essentially implies the increase of flex-
ally designed model Beam HB4-10L-T65 and HB3-10L- ural and shear crack widths, thus the reduction of aggregate
T50, respectively. Note that model Beam HB3-12L-T50 interlock mechanisms. Therefore, the limited elongation in
sustained a maximum average shear stress of Vmax /bd = the beams with the vertically distributed flexural reinforce-
0.65 f c ′ MPa (7.8 f c ′ psi). ment indicates improved resisting mechanisms to effectively
delay or avoid premature sliding shear failure.
Effects of flexural and shear reinforcement
contents SUMMARY AND CONCLUSIONS
Although all six beams developed flexural yielding as a Six high-strength concrete short beams with length-to-
consequence of following the capacity design approach depth ratios of 3.0 to 4.0 (shear-span ratios of 1.5 to 2.0) and
based on the current code, the ultimate ductility levels flexural reinforcement ratios of 2.3 to 4.5 percent (tensile
achieved by the model beams were different. The spacing of steel ratios 1.15 to 2.25 percent) were experimentally tested
stirrups in the model beams reinforced with six No. 19 longi- under combined cyclic shear and double curvature bending.
tudinal bars was 100 mm center-to-center, which was The following conclusions can be reached based on the
provided following the maximum transverse reinforcement experimental results.
spacing limit d/4 recommended by the seismic design 1. Flexural yielding can be developed in short HSC beams
provisions of the ACI 318-95 Code.8 The actual shear with length-to-depth ratios of 3.0 to 4.0 (shear-span ratio of
strengths of these beams were higher than the calculated shear 1.5 to 2.0) by using the capacity design approach based on
demands. Satisfactory response and an ultimate ductility the seismic design provisions of the current ACI 318 Code.
factor of about 6.0 were achieved by these model beams. However, such design could not prevent the eventual sliding
On the other hand, the heavily reinforced short beams shear failure of the beams corresponding to a ductility factor
were designed with their shear reinforcement providing a of 3.5 to 6.2.
shear strength essentially equal to the maximum probable 2. The equivalent rectangular stress block approach of the
shear demand. As shown in Fig. 5 and 7, the heavily rein- current ACI 318 Code can provide predictable but somewhat
forced conventional Beam HB4-10L-T65 and HB3-10L-T50 conservative estimates of the flexural strengths of the short
developed limited ductility levels of 4.0 and 3.5, respec- HSC beams.
tively, showing a reduction in ductility compared with 3. The ductility of HSC beams of conventional design
lightly reinforced beams. Such a trend of reduction in decreases with increasing flexural reinforcement, as a conse-
ductility with the increase of flexural reinforcement is effec- quence of increased shear demand.
tively improved by using vertically distributed longitudinal 4. For heavily reinforced HSC short beams, the seismic
reinforcement, as demonstrated by the improved perfor- behavior can be effectively improved with increased
mances of HSC Beam HB4-12L-T65 and HB3-12L-T50. ductility by distributing the flexural reinforcement vertically
in the beam section.
Beam elongation responses 5. The vertically distributed flexural reinforcement can
The improved resisting mechanisms of beams with distrib- effectively restrict the beam elongation, thus maintaining the
uted flexural reinforcements can be further evaluated by aggregate interlock mechanisms, and as a consequence,
studying their longitudinal responses. Fig. 8(a) and (b) show delaying the occurrence of sliding shear failure due to cyclic
the longitudinal elongation versus lateral deformation loading.
hysteretic relationships for beams with conventional rein-
forcement (dashed curves) and beams with vertically distrib- ACKNOWLEDGMENTS
uted flexural reinforcement (solid curves). In Fig. 8, the The research described in this paper has been funded by Carpen-
ters/Contractors Cooperation Committee and conducted in close collabora-
horizontal axes show the lateral displacement drift ratio, and tion with engineers from Englekirk and Sabol, Inc. Valuable input from S.
the vertical axes show the average longitudinal elongation D. Nakaki and R. Englekirk is warmly appreciated. The research project
strains along the beam end portions where sliding shear also received support from the Concrete Reinforcing Steel Institute (CRSI),
failure took place. the Portland Cement Association (PCA), Fontana Steel, Inc., and the
Southern California Carpentry Joint Apprenticeship and Training
As shown in Fig. 8, the longitudinal elongation deforma- Committee. The authors would also like to use this opportunity to pay their
tions are approximately symmetric in the push and pull tribute and respect to J. H. Nevin, late manager of the Western Regional
lateral loading directions. In all cases, the elongation is small Office of CRSI, for his valuable contributions to the concrete industry and
during loading cycles corresponding to small peak lateral education.
displacements. There is an apparent increase of elongation
after the beams developed flexural yielding, indicating the NOTATIONS
Av = cross-sectional area of stirrups within spacing s
significant shift of neutral axis and reduction of compression d = effective depth of beam
zone in the beam sections. Under cyclic loading, the elonga- fc′ = concrete compressive strength
tion deformation of a flexural reinforcing bar yielded in fy = yield strength of reinforcement
tension cannot be fully recovered due to the compression in L = clear span length of beam
the subsequent reversed loading. As a consequence, the elon- Mif = ideal moment strength of beam
Mn = nominal moment strength of beam
gation of the beam accumulates upon the application of Mpr = probable ultimate moment strength of beam
loading cycles corresponding to larger displacements, as s = stirrup spacing
exhibited in Fig. 8. Vif = estimated ideal flexural strength

ACI Structural Journal/May-June 1999 399

Vmax = experimental maximum shear force 8. ACI Committee 318, “Building Code Requirements for Reinforced
Vn = nominal shear strength Concrete (ACI 318-95) and Commentary (ACI 318R-95),” American
Vs = shear strength contributed by stirrups Concrete Institute, Farmington Hills, Mich., 1995, 369 pp.
Vu = required shear strength 9. Tan, K. H. et al., “High-Strength Concrete Deep Beams with Effective
Vy = shear force at first yield of flexural reinforcement Span and Shear-Span Variations,” ACI Structural Journal, V. 92, No. 4,
Δ1 = average measured displacement for push and pull directions at Vy
July-Aug. 1995, pp. 395-405.
Δyi = yield displacement
μΔ = displacement ductility factor 10. Paulay, T., “Simulated Seismic Loading of Spandrel Beams,”
Journal of the Structural Division, ASCE, V. 97, No. ST9, Sept. 1971,
pp. 2407-2419.
REFERENCES
1. Shah, S. P., and Ahmad, S. H., eds., High Performance Concrete: 11. Paulay, T., “Coupling Beams of Reinforced Concrete Shearwalls,”
Properties and Applications, McGraw Hill, 1993, 403 pp. Journal of the Structural Division, ASCE, V. 97, No. ST3, Mar. 1971,
2. Hester, W. T., ed., High-Strength Concrete—Second International pp. 843-862.
Symposium, SP-121, American Concrete Institute, Farmington Hills, 12. Paulay, T., and Binney, J. R., “Diagonally Reinforced Coupling
Mich., 1990, 786 pp. Beams of Shearwalls,” Shear in Reinforced Concrete, SP-42, American
3. Shin, S. W.; Ghosh, S. K.; and Moreno, J., “Flexural Ductility of Concrete Institute, Farmington Hills, Mich., 1974, pp. 579-598.
Ultra-High-Strength Concrete Members,” ACI Structural Journal, V. 86, 13. Theodosios, P.; Tassios, M. M.; and Bezas, A., “On the Behavior and
No. 4, July-Aug. 1989, pp. 394-400.
Ductility of Reinforced Concrete Coupling Beams of Shearwalls,” ACI
4. Fang, I. K.; Wang, C. S.; and Hong, K. L., “Cyclic Behavior of High-
Strength Concrete Short Beams with Lower Amount of Flexural Reinforce- Structural Journal, V. 93, No. 6, Nov.-Dec. 1996, pp. 711-720.
ment,” ACI Structural Journal, V. 91, No. 1, Jan.-Feb. 1994, pp. 10-18. 14. Paulay, T., and Priestley, M. J. N., Seismic Design of Reinforced
5. Sugano, S. et al., Experimental Studies on Seismic Behavior of Rein- Concrete and Masonry Buildings, John Wiley and Sons, Inc., 1992, 744 pp.
forced Concrete Members of High-Strength Concrete, SP-121, W. T. 15. Wong, P. K. C.; Priestley, M. J. N.; and Park, R., “Seismic Resistance
Hester, ed., American Concrete Institute, Farmington Hills, Mich., 1990, of Frames with Vertically Distributed Longitudinal Reinforcement in
pp. 61-87. Beams,” ACI Structural Journal, V. 87, No. 4, July-Aug. 1990, pp. 488-498.
6. Sakaguchi, N. et al., “Shear Strength of High-Strength Concrete
16. Priestley, M. J. N., “Myths and Fallacies in Earthquake Engi-
Members,” High-Strength Concrete—Second International Symposium,
neering—Conflicts between Design and Reality,” Concrete International,
SP-121, W. T. Hester, ed., American Concrete Institute, Farmington Hills,
Mich., 1990, pp. 155-178. V. 19, No. 2, Feb. 1997, pp. 54-63.
7. Xiao, Y., and Ma, R., “Seismic Behavior of High-Strength Concrete 17. CRSI, “Using Soft Metric Reinforcing Bars in Non-Metric Construc-
Beams,” International Journal of the Structural Design of Tall Buildings, tion Projects,” Engineering Data Report, Concrete Reinforcing Steel Insti-
Mar. 1998. tute, No. 42, 1997, Schaumburg, Ill.

400 ACI Structural Journal/May-June 1999