Xe RSs elente svar eleainy TECHNICAL PAPER \
Title no, 99-857
Influence of Beam Size, Longitudinal Reinforcement, and
Stirrup Effectiveness on Concrete Shear Strength
by Eric J. Tompos and Robert J. Frosch
‘Research has shown thatthe current ACI shear design provisions
provide unconservative results for large beams and beams with
Tow levels of longitudinal reinforcement. Furthermore, recent
research has indicated that some assumptions made in the current
design provisions for shear reinforcement may further reduce the
level of conservatism. To farther explore these deficiencies, this
study investigated the influence of beam size and longitudinal
reinforcement ratio on the shear strength attributed tothe concrete,
{aswell as the effectiveness of stirups in transferring shear across
{a diagonal erack. The experimental portion ofthis study tested six.
rectangular reinforced concrete beams. Other research data found
in th literature were used t0 supplement the experimental data
Based on test results and a data analysis, conclusions regarding
the influence of beam sie, longitudinal reinforcement, and transverse
reinforcement on shear strength are presented herein.
‘Keywords: beam; longitudinal reinforcement; sear; strap
INTRODUCTION
Although the shear design provisions prescribed by
‘ACI 318-99! have been used successfully for years, research
Suggests these provisions may provide unconservative results
‘when applied to large beams and beams with low longitu-
dinal reinforcement ratios p. Under these conditions, it has
been shown? that the shear strength attributed to the concrete
decreases as either beam depth increases or p decreases. In
combination, these two effects can result in shear strengths
significantly lower than those calculated using the provisions
of ACI 318.99.
‘Adding minimum shear reinforcement enhances shear
resistance; however, recent research* has shown that the
shear strength attributed to stirrups may also be adversely
affected in large beams, Factors ascribed to this strength
reduction include: variations in shear crack inclination, stirrup
anchorage, and discretization of stirrup strength.
‘This study futher investigated the influence of the previously
mentioned factors on the shear strength attributed to both the
concrete and stirrups as stipulated by the current ACI shear
design provisions. Additionally, the data used to establish the
‘current design provisions for concrete shear strength are assessed
in terms of whether they provide an accurate representation of
shear behavior in light of more recent studies. Ths investigation
clearly illustrates that both concrete and stiup shear strength can
be considerably lower than commonly assured in design.
RESEARCH OBJECTIVE
The research objective was to further investigate the effects
‘of beam size and stirrup effectiveness on the shear strength of
reinforced conerete beams.° In particular, this study focused
‘on the effects of beam size and longitudinal reinforcement ratio
regarding the shear strength attributed to the concrete. Further-
‘more, this study investigated several factors that influence
ACI Structural Journal/September-October 2002
series
=
28 _
17 ele
iif | Series 1
e 1
rit |
bead |
| ee
jt Heekt. J]
- oe 4
y362
Vie3 :
Fig. J—Beam cross sections.
stirup effectiveness. These factors include: distribution of
transverse reinforcement, stimup anchorage, discretization of
stirup suength, and shear crack inclination
EXPERIMENTAL INVESTIGATION
Six beams were tested in an experimental investigation
conducted at Purdue University.® The two primary variables
influencing the specimen design were the beam depth and the
stimup spacing. Otber parameters, such as the shear spanto-
depth ratio ald and te longitudinal reinforcement rao, were
kept constant. The experimental progrem was divided into wo
separate test series: Series I and IL
Series I was comprised of two beams designed according to
‘ACI 318-99, An overall beam depth of 36 in. was chosen be-
cause it approaches the maximum depth permitted by the ACL
Code without requiring longitudinal skin reinforcement
(Fig. 1). The inclusion of skin einforcement was undesirable
because ofits potential shear strength contribution. Finally, the
design of the Series I beams was consistent with beams tst-
‘ed previously’ to allow for comparison.
Series Il was comprised of four beams designed as balf-scale
models ofthe Series I specimens. This significant change in size
was chosen to accentuate any effecsatribuable to beam size.
‘When scaling the specimens, the effective depth, beam width,
and shear reinforcement spacing were scaled down by one-half.
Toraintain compliance withthe ACI Code, however, the clear
cover was not scaled. Scaling inthis manner resulted inthe Se-
Fes I specimens shown in Fig. 1
‘ACI Scan 99. No.5, Sepeber Octobe 202
IR at 3 are ee te
a allt eoper nes peaon sted fom he sopra per
Rien lc haps Ul
559
0"
5{ACI manr Ene J. Tops avenger wih NIN. nc, Napanee, nd. He
recied hs BSCE and MSCE from Purse Uniraty Wet fe Ind.
‘AC member Robert J. Frosh so seit peor of lengli Park
Unies erected hs BSE fom Tare Universi, New Oo id
[MSE ond PRD fm he Unies of Ts ata, Tex. Scary af
{ACI Conmiee 22, Cracking. ond & member of ACI Conmanees 1%. Sey
‘Serica, ad nay 408 Bond nd Denopment of Rrrcement a
Fer Reoced Poymer Reinforcement.
Table 1—Stirrup layout
seies Specimen] east 0m | ax |b.
ol@fel o fTolel o
1 Lez 2 [612 Toone [55 | ao"
vies [3 [ese | ose [a7 [oh
visa [=| = [oa | oan | — 0
y [wea eae fone Pas |
wie2e| 27916 [ove [aa9 |
vies [3 [77s [03st | aay | a0
“Minium ping bse on AC 31899 eaten 11.553, By, 1-1 Clan 7
‘atedon nomial fee eng fi
Table 2—Reinforcing steel strength
Yall, aT
Design roan] “En on
oto a
Sarai —[ 8 0
Sia laimge [3 [8 iis
lonphnalbas | 8 % t01
‘Series If Dywidag bars ® 132 13
Series I longitudinal bars| os | 70. 4105 -
a
Table 3—Concrete mixture design and properties
Teiesl | — Seder
ow che a)
A oo Sieve | Om
i 2 ‘34 in, coarse aggregate (SSD) 1850 Ibiyd? | 1850 tbvyd?
0 Fine aarewe (SSD) | “tzzz mys? | 1356 hy
» Weuye | Toye
0 Bony’ 9oulyd®
0 2000 psi__| 00 pe
me om 00m ous eom 0010 oon ow ome eo | at
Sein a) eo
Fig. 3No. 2 reinforcement stress-strain curve.
Forall six specimens, p was kept constant at 1%, This amount
of flexural reinforcement was selected because it represents a
‘common design condition. As for shear reinforcement, the
Series T beams were designed to investigate the ACI required
minimum shear reinforcement strength. Accordingly, the
Series I shear reinforcement was designed to provide a shear
stress of 50 psi (Ay min = S0by.ff;). Using a nominal yield
strength of 60 ksi, stizmups fabricated from No.2 and 3 deformed
reinforcing bars were spaced as shown in Table 1.
Series Il beams were constructed both with and without
shear reinforcement. In beams with shear reinforcement, the
stirrup spacing was scaled to provide the same dis as that of,
Specimen V36-3. One beam containing stirrups, Specimen
V18-2c, was constructed with closed stirrups to investigate
whether closed stirups are more effective than U-shaped
stirrups. Finally, Series HI included a beam without shear
reinforcement, Specimen 18-0. This beam was constructed to
investigate whether the concrete shear strength is affected by
the presence of shear reinforcement. A summary ofthe variables
studied in this experimental program is provided in Table 1
560
| 680 pat
For both test series, the loading arrangement consisted of a
concentrated load applied at midspan as shown in Fig, 2.
Loading in this manner produces @ constant applied shear
throughout both shear spans, as well as a linearly varying
‘moment that reaches a maximum under the applied load. The
Tength of the shear span was selected based on past research,”
which has shown that beams with alds greater than 2.5 resist
shear primarily though beam action rather than arch action,
Consequently, an a/d of 3.0 was chosen to minimize the
shear strength contribution of arch action, thereby providing.
lower-bound test results. Using an ald of 3.0 the clear spans
were set at 16 ft, 9 in. for Series I and 8 fy, 4.5 in, for Series I,
as shown in Fig. 2.
MATERIALS
Reinforcing steel
Longitudinal reinforcement consisted of three bar types:
No. 8 Grade 60, No. 9 Grade 60, and | in.-diameter Grade 150
threadbar. To ensure uniform chemical and mechanical proper-
ties, bars ofa given size were obtained from the same heat of
steel. The results of tensile tests performed on representative
ACI Structural Journal/September-October 2002= L
Lo sear cn Fron
3
or 0203 ad OS OKO
Miepn Detection in)
Fig. 4—Load versus deflection (Specimen V36-3).
‘steel coupons are presented in Table 2. Itshould be noted that the
hotrolled, deformed No. 2 reinforcing bars did not exhibit a
‘well-defined yield point, as shown in Fig, 3; therefore, the yield
stress was determined using the 0.2% offset method,
Concrete
Al beams of a given test series were cast from a single
batch of conerete. A nominal 4000 psi concrete mixture was.
used for Series I. Atthe time of testing, this mixture achieved.
‘an average compressive strength of 6200 psi. A similar concrete
‘mixture was used for Series I however, this batch of concrete
achieved an average strength of 5200 psi at the time of testing.
‘Table 3 provides the details ofthe concrete mixture design and
mechanical properties.
‘TEST SETUP AND PROCEDURE
Instrumentation was provided to measure the applied load,
‘beam deflection, and strain in the reinforcement. During con-
struction of the reinforcing cages, electrical resistance strain
‘gages (0.125 in. gage length for stimups, 0.250 in. gage length
for longitudinal bas) were placed on stirrups that the inclined
crack was expected to cross, These gages were positioned
‘vertically on the stirrups to place them as close as possible to
the anticipated crack. Linear variable displacement transducers
(LVDTs) were placed to measure beam deflections at the
supports, quarter spans, and midspan. A 600 kip universal
testing machine monotonically applied the concentrated load
at midspan while load cells monitored the applied load. Where
concentrated loads were applied, 6 in.wide steel bearing
plates were used to prevent local crushing of the concrete.
Round steel bars placed between the 6 in.-wide plates approx-
mated simply supported conditions.
STRUCTURAL BEHAVIOR
Series |
In both specimens, flexural cracking began at approximately
the same applied load. As the load was increased further, new
flexural cracks formed in the shear span. These flexural cracks
increased in depth and began to incline towards the applied load,
‘becoming flexural shear cracks. The primary shear crack formed.
at approximately 74 and 71 kips in Specimens V36-2 and V36-3,
respectively (Table 4, The formation ofthe primary shear crack.
is evident in the Ioad-deflection plot of Specimen V36-3, as can
be seen in the shift in the curve of Fig, 4, The formation of the
primary shear crack was also observed in the readings taken
from the strain gages on the stirrups.
ACI Structural Journal/September-October 2002
UM
Specimen 26:2
e fae
Fig. 5—Series I crack patterns.
00 a 02 0 on os
Mapa Defcon in)
Fig, 6—Load versus deflection (Specimen V18-2).
For both specimens, failure occurred when the primary
shear erack propagated deeply into the compression zone,
thereby reducing the compression zone to the point of buckling
outward under combination of shear and compressive stress.
The primary difference between the observed cracking
patterns was the angle at which the primary shear crack
formed. As shown in Fig. 5, the crack inclination observed
in Specimen V36-3 was much steeper than that observed in
Specimen V36-2.
Series it
For the Series II beams, a representative load-a2] — | — | —
Series avenge] 196
Note Colima &¥,.
date from other studies in the literature. This data survey
included the data used to formulate the current ACI shear
design provisions.
‘The current ACI provisions related to the shear strength
attributed to the concrete were developed based on test
results from 194 rectangular beams,*""” which are plotted in
Fig. 8. Although this figure appears to validate the use of
2,Jf2 in shear design, the 194 test results provide a poor
representation of overall shear behavior due to high reinforee-
‘ment ratios and relatively small beam sizes. It should be noted
that, while 1% longitudindl reinforcement is common in
design, fewer than 10% of the original test data were from
beams with 1% longitudinal reinforcement or less, Further-
‘more, ofthe beams with 1% or less longitudinal reinforcement,
only one failed ta shear stress below 2(f.’. This facts likely
due to the limited range of beam depths investigated in the
original test data. As the histogram in Fig. 9 shows, 87% of
the beams considered had effective depths less than 15 in
The data from this study indicate that an effective depth of
16-3/4 in, (Series ID) isnot large enough to exhibit a strength
reduction due to beam size. In summary, a combination of
high reinforcement ratios and small beam sizes likely resulted
inhigh shear strengths in the data used to formulate the current
shear design provisions, thereby hiding the true behavior of
large and lightly reinforced beams.
Recent research
To address the apparent limitations in the data used to
develop the current design provisions, a review of more
ACI Structural Journal/September-October 2002Fig. 10—Concrete shear strength versus longitudinal
reinforcement ratio (recent data).
recent research, was performed. Test data were obtained
from 18 sources >>"!?8 These data were from tests of simply
supported beams of constant rectangular cross section, incl
ing beams subjected to concentrated or distributed loading, and
beams with and without shear reinforeement, The data were
reduced by limiting concrete compressive strength to 8000 psi
Orless and by excluding results from beams with afd es than
25 to eliminate data from beams potentially transferring
shear through arch action. Additionally. all specimens included
‘were reported to have failed in shear."The resulting dataset,
plotted in Fig. 10, consists of 414 test results. Inthe plot
the data are partitioned by the effective depth tothe tensile
reinforcement to illustrat the effect of beam size.
Several observations can be made from Fig. 10, the
ost apparent being the adverse affect thatthe longitudinal
reinforcement ratio has on beams with 1% longitudinal
reinforcement or less, At 1% reinforcement, beams ofall.
effective depths fail at shear strengths below 2 fbyd
AAs p decreases below 1% longitudinal reinforcement, the
concrete shear strengths for beams ofall sizes decrease
further. Figure 10 also indicates thatthe detrimental effect
of low longitudinal reinforcement ratios is exacerbated
by beam size. Overall, larger beams, which are shown as
triangle data points in Fig. 10, exhibited the lowest shear
strengths for a given reinforcement ratio. Infact, the lowest
shear strengths Were observed in the beams with the
greatest effective depths. These beams, which had effec-
tive depths of 79 and 118 in, exhibited concrete shear
strengths below 1.0,f,’byd Finally itis important to note
that, while the size effect is observed in Fig. 10, a large
amount of scatter exists in the dats, Considering this seat
ter, it seems that the longitudinal reinforcement ratio
plays a more significant role in influencing the concrete
shear strength than does the size effect.
SHEAR REINFORCEMENT
Stimups provide tle, if any, contribution to shear strength
until crossed by a shear crack, making it important to under-
stand shear cracking behavior and its effect on stimup
strength. Figure 11 depicts the primary shear cracks that
formed in the specimens of the present study. As shown,
the angle at which the primary shear crack formed varied
between 31 and 54 degrees. For comparison, the diagonal
dashed lines in Fig, 11 correspond to the crack inclination
implied by the ACI Code. This crack inclination, based on dis,
is slightly less than 45 degrees from horizontal
ACI Structural Journal/September-October 2002
Specimen V18-2 & V18-2c
Fig. 11—Primary shear crack comparison.
‘As shown in Fig. 11, the V36-2 crack is shallower than the
‘ACT assumed crack angle, while the V36-3 crack formed at
fan angle slightly steeper than that assumed by ACI. This
‘observation is significant because the angle at which the
‘crack forms affects the stirrup shear strength. A shallow
crack, similar to that observed in Specimen V36-2, has a
greater horizontal projection and crosses more stirrups than
a steeper crack, as in Specimen V36-3. Furthermore,
comparison of crack inclinations between the full- and helf-
scale beams, illustrated in Fig. 11, indicates that beam depth
does not significantly influence the shear crack formation
angle. This variation in shear crack angle can significantly
influence the shear strength provided by shear reinforcement
‘and requires additional consideration.
Crack angle and stirrup anchorage
In ACT 318-99, dis is used to compute the number of
rups crossing a shear crack forming at an assumed angle of ap-
proximately 45 degrees. This ratio, as used in the ACI Code,
seldom results in whole numbers, allowing for fractional
stirrup contributions to shear strength. In actuality, a crack
‘cannot cross a fractional portion ofa stirup. Intuitively, if a
sirup is crossed by the shear crack, it contributes strength;
iffa stirrup is not crossed by a shear crack, it should not con-
tribute strength. This principle is easily incorporated when
calculating shear strength by truncating the decimal portion
of dfs, thus producing an integer quantity. While this method
provides a more realistic estimate of the number of stirrups
crossed by an inclined crack, it does not provide a lower-
563‘Table 6—Tests by Johnson and Ramirez"?
Specimen V2
Fig. 12—Primary shear crack comparison from Reference 4
‘Table 5—Shear crack formation
sMieta, | Shear’ | stew, | enoaine | oe,
Specimen | “tips” [Yow te] hips" [stirpa Ny| NICS)
(a
vae2_| i096 [18 | a6] — 7] a8
vies [uso] | a0 [ 2] 343
visa | 3867_| a4] 3] 2 | 38
Viz [34a [200 | 4a [2 is
ve _| 7 [32] ass] 1 [asa
2
bound shear strength unless all stirups crossed by a shear
crack are fully effective
Figure 1 superimposes the locations ofthe primary shear
cracks with the stiups. As shown in the figure, shear cracks
Propagate very high into the section as they extend toward
the
lied load. When stirrups are crossed by a shear crack
high in the section, they often have a small development
length and may not reach yield, as assumed in design calcula-
tions. Although these stirrups are capable of contributing
some shear strength and could be accounted for through the
use of fractional stirrups, the level of stress developed in
these stirrups cannot be accurately estimated. By neglecting
the contribution of these inadequately developed stirrups,
however, a lower-bound shear strength can be computed. TO
identify ineffective stirrups, the development length that re-
sults when a shear crack crosses a stirrup can be compared with
the required development length. When the required develop-
ment length is not provided, the sirup is assumed to be inef-
fective. ILis important to note that, due to the continuity atthe
‘bottom ofthe U-shaped stirrup, the stirrups were assumed to
be fully effective atthe bottom of the beam.
‘Stirup development lengths can be approximated using
Eq. (1), which is a modified version of the ACI equation
564
Cracking
Concrete |Appliodshea| shear Von | Stinup shear
Specimen _|prengthfo.ps] Vankins | kips | Ve kis
w, @ o @ S
2 3280) 30 2 5
3 10,907 8 3 ®
4 10,390 7 32 is
7 7420 6 8 18
3 7440 58 s i
‘ac: ca einud nad ott Ges wed Cauon 3 V,=Var~ Vor
(Section 12.5, ACI 318-99) for calculating the development
length of hooked bars.
ly = bd fy w
‘TB I @.00
The original ACI expression computes development length
‘based on the reinforcing bars developing 1.25f, to ensure
ductility. Because the actual yield strengths are known in this
‘case, Eq. (I) is divided by 1.25 to remove this overstrength
factor. Using the actual material properties for the present
study, development lengths of 4.0 and 5.9 in. were computed
for the No. 2 and 3 reinforcing bars, respectively. These
computed development lengths are shown as horizontal
dashed lines in Fig. 11. From this figure, the number of fully
effective stirrups can be determined by counting the stirrups
that are crossed by the shear crack below the required develop-
ment length. The number of effective stirrups estimated in this
manner is presented in Table 5.
Reviewing the results for Specimen V36-2, the crack pattern
shown in Fig, 5 indicates seven effective stirrups; however, the
‘experimental results in Table 5 suggest that only five stirrups
‘contributed strength. This discrepancy may be the result of
Bq, (1) inaccurately calculating the development length for
the No. 2 reinforcing bars. In Specimen V36-3, two stirrups
were effective, but the. test resulted in a slightly greater
strength, which is likely to be the result of strain hardening
in the stirrups near ultimate,
‘The Series Il test results, also shown in Table 5, compare
‘well with the calculated strengths. It should be noted that
‘Specimens V18-2 and V18-2c were identical in construction,
‘except that Specimen V18-2 had U-shaped stirrups and
‘Specimen V18-2c had closed stirrups. Comparing the ultimate
loads in Table 5, Specimen V18-2 sustained slightly more
oad than Specimen V18-2c, More importantly, however, is
the considerable difference in the inclination of the primary
shear crack, as shown in Fig. 11.
‘Specimen V 18-2 cracked at an angle shallower than that as-
sumed by ACI, thereby crossing two stirups (Stirrups 3 and 4
in Fig. 11) below their development length, making them fully
effective. This observation is reflected in the ultimate
strength of Specimen V18-2, which also indicates two effec-
tive stirrups. Also shown in Fig, 1 is the primary shear crack
in Specimen V18-2c, which formed at an angle steeper than
that assumed by ACI. If U-shaped stirrups had been used in
‘Specimen V18-2c, it is possible that only one stirrup would
have been fully effective. The closed stirrups in Specimen
VI8-2c, however, were estimated to require significantly ess
length to develop than did U-shaped stirrups. This reduced de-
velopment length was estimated to be 2 in.,* and is shown in
ACI Structural Journal/September-October 2002i
Table 7—Analysis of stirrup capacity
]Epesnennl | Dinwete | Ea. ¥a Bperinenal peel
specimen | Vetips | atmpst | “iipe” [ActVakipe] Ea Y, || ACI,
@ @ @ @ @ @ w
Vael_| 356 < ma [ asa 12 uo
vaes_| 0 1 Ta | 394 26 tI
Vie Tz T 5 157 35; HI
views |i z be [is To a
vi 155 o sa] 354 10 a
ve 25a T isa [35a a
2 8 t 69 Bs 12 as
zl 5 i 2 16 12 06
a @ i 69 Bs 28 a
= a 169 136 a 3
¢ 5 ps 16 19 to
eng isa oa
Fig. 11. The shear crack crosses Stirrup 5 below this re-
duced length, providing adequate development. As aresult,
the shear crack in Specimen VI8-2c erosses two stirrups below
their development lengths, resulting in an ultimate strength
very close to that of Specimen V18-2.
Based on the comparison of only the ultimate loads, a compar-
ison between Specimen V18-2 and V18-2c does not illustrate
‘benefit from the use of closed stirrups. From this analysis,
however, itcan be seen that closed stirrups may mitigate the
strength reduction that occurs due to variations in shear
crack inclination. This is attributed to the smaller develop-
‘ment length of the vertical stirrup leg and its relationship to
the shear crack location.
‘Although the Series I results did not conclusively show a
relationship between overall stirrup strength and the number
Of stirrups crossing the shear crack, the experimental results,
cof two previous studies*'® demonstrate this relationship and
its importance. In one study,* two beams identical in design
to Specimen V36-3 were tested. These beams were both cast
from the same batch of concrete and reinforced with steel
from the same heat. In this investigation, the shear reinforce-
‘ment yield strength was 70 ksi and the concrete compressive
strength at the time of testing averaged 5300 psi. The test
setup and procedure were identical to those of the present
study. Test results for these beams, Specimens VI and V2,
are presented in Table 5, and the primary shear cracks are
shown in Fig. 12
‘A study conducted by Johnson and Ramirez"? tested five
beams that are two-thirds the size of the Series I specimens.
‘These 12x 14 in, beams contained 2.4% longitudinal reinforce-
ment placed at an effective depth of 21.2 in. Each shear span
‘was 65.8 in. long, resulting in an a/d of 3.1. Additionally,
these five beams contained shear reinforcement, which provided
the ACI minimum shear reinforcement strength of 50 psi. The
shear reinforcement consisted of No. 2 deformed reinforcing.
bars similar to those used in the present study. The resulting.
stirrups were spaced at 10.5 in., giving a dis of 2.0. These
bars did not exhibit a well-defined yield stress; however, ata
26% strain offset, the yield stress was found tobe approximately
69.5 ksi. The test results for five of the beams from this study
are presented in Table 6. It should be noted that these specimens,
are identical to each other in design, differing only in concrete
compressive strength as shown in Table 6
‘The procedure for assessing stirrup effectiveness, as demon-
strated previously, was performed on the data from the previous,
ACI Structural Journal/September-October 2002
studies. As shown in Table 7, several specimens from the
previous studies—Specimens V1, 2, and 3—failed at shear
strengths considerably below the strength calculated by
ACI318-99. Infact, these specimens exhibited shear reinforce-
‘ment strengths as fow as 40% of the calculated values. The
shear crack angle in the low-strength specimens was steeper
than that assumed by ACI, and was considerably steeper than
those observed in companion specimens failing at greater
strengths (Fig. 12). Based on these results, it appears that
fewer stirrups were fully developed in these specimens due
toa steep shear crack inclination, thereby resulting in lower
shear strengths.
inforcement distribution
The distribution of shear reinforcement can significantly
impact the overall shear strength of a reinforced concrete
‘beam. For example, if inclined cracking occurs ata steep angle
and one stirrup is tendered ineffective, a beam with widely
spaced stirrups—such as Specimens V36-3, V1, V2,2, and 3—
may experience up to a 50% reduction in the strength provided
by the stirrups. Alternately, for beams with enhanced stirrup
distributions, such as Specimen V36-2, the effectiveness of @
single stirrup becomes less important. The stirrup strength is
only reduced by 20% if one stirup is rendered ineffective.
‘Specimen V36-2 was constructed forthe purpose of inves:
tigating the influence of shear reinforcement distribution. In
this study, d/s was used to assess the stirrup distribution.
Specimen V36-2 had a dis of 5.2, while the other 36 in-deep
specimens—Specimens V36-3, V1, and V2—had ds ratios
of 2.3. Although the dis ratio differs all beams have the same
design capacity (nominal strength shear V,). This similarity
in strength can be seen in Table 7, where the actual stirrup
shear strength of Specimens V36-2, V36-3, and V2 (Expeni-
‘mental V,) is approximately equal considering the difference in
stirup yield strengths. Conversely, Specimen V1 failed with
‘an average stirrup strength that was approximately 1/2 the
strength of the companion specimens and 1/2 the value ce
culated by ACT. Similar results were observed in the Johnson
and Ramirez'” study, where Specimens 2 and 3 exhibited @
shear strength approximately one-half that of the companion
specimens (Specimens 4, 7, and 8). As mentioned previously,
this reduction in stirrup strength i possibly the result of only
‘one stirrup remaining fully effective after the formation of
the primary shear crack. With a tighter stirrup spacing, this
strength reduction may have been mitigated. These experimental
565results illustrate the significance of transveise reinforcement
distribution and the benefit of considering reinforcement dis-
tribution in design,
Calculation procedure
‘The method for assessing effective stirrups used herein
Was incorporated into a calculation procedure in Reference 4
and is presented here as Eq. (2). This procedure utilizes an
equation similar to the current ACI stirrup design equation,
but instead of using a/d to compute the number of effective
stirrups, a new term Nis provided for this purpose. This term
isa whole number and accounts for stirrup development.
Ne ®
area of shear reinforcement;
yield strength of shear reinforcement;
number of stimups that effectively transfer shear
across the inclined crack (integer quantity);
tent
distance fom extreme compress
of longitudinal reinforcement;
gy = development length of shear reinforcement (refer to
Bq, (1); and
spacing of sheer reinforcement.
Equation (2) was evaluated using the actual material
strengths forthe beams shown in Table 7. In these calculations,
the stimup development lengths were estimated at 5.8 and
4.0;n. for the beams from References 4 and 19, respectively
For comparison, the shear strength attributed tothe stiups V,,
calculated using the current ACI expression, is also provided
in Table 7.
Overall, Equation (2) provides conservative estimates of
stirrup shear strength for the beams presented in Table 7.
‘The computed strengths for six of the 12 beams were very
conservative, with the strengths estimated using Eq. (2)
accounting for less than one-half of the experimentally ob-
served strengths. In these instances, the primary shear crack
formed at shallow angles, resulting in more effective stirrups
than estimated by Eq, (2). Contrary to this, where the prima-
1 shear crack formed at a steep angle—as in Specimens V1,
VI8-26, 2, and 3—Eq, (2) provides a reasonable lower
‘bound for stimup strength, whereas the ACI equation is very
unconservative, Since the angle of shear crack formation
cannot be predicted, calculation ofthe lower-bound strength
using Eq, (2) seems appropriate
fiber to centroid
‘SUMMARY AND CONCLUSIONS
Six beams were tested to investigate the effects of beam size
and stirrup effectiveness on the shear strength of reinforced
concrete beams. The test configuration consisted of simple
supports with a concentrated load applied monotonically at
midspan, thereby producing a constant applied shear
throughout the shear span. To investigate the effects of beam
size on shear strength, beams of two different sizes were tested.
While the beam size was varied, the longitudinal reinforce-
‘ment ratio was held constant. In the present study, shear
reinforcement was included in five of the six beams to investi-
Bate various aspects of stirup effectiveness. The test variables
included stirrup spacing and stirrup anchorage details. In
addition to an experimental investigation, a survey of data in
the literature was performed to gain insight into the influence
566
of longitudinal reinforcement on shear strength. This data
review included the data used to derive the current ACI shear