ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 106-S39

Shear Strength Evaluation of Concrete Bridge Girders

by Afshin Esfandiari and Perry Adebar

A shear strength evaluation procedure for structural concrete such programs to check numerous sections along a bridge.
girders that contain at least minimum transverse reinforcement Additionally, the program options may result in different
(stirrups) is presented. The procedure is similar to the 2008 users reaching different conclusions for the same girder, and
AASHTO LRFD shear design method except that it does not it is not possible to confirm results of such programs using
require trial-and-error for shear strength evaluation and provides
hand calculations.
more insight by providing information about different shear failure
modes involving stirrups yielding, diagonal crushing of concrete,
and longitudinal reinforcement yielding. To understand the RESEARCH SIGNIFICANCE
background, predicted trends are compared with the modified This paper presents a shear strength evaluation procedure
compression field theory (MCFT) for the idealized case of an that gives considerable insight into the shear failure mode of
element subjected to uniform shear, and are compared to concrete bridge girders and yet is simple enough that a user
Response-2000 for beam elements subjected to combined shear can implement the procedure into a small computer program
and bending moment. To validate the proposed method, shear strength for checking numerous sections along a bridge, and can
predictions are compared with results from strength tests on reinforced confirm the results of the computer program by hand calcula-
and prestressed concrete beams, as well as predictions from shear tions. To validate the procedure, shear strength predictions are
design methods in AASHTO LRFD and ACI 318. The procedure is used
compared with results from strength tests on reinforced and
to evaluate the shear strength of girders in three existing bridges with
prestressed concrete I-girders, prestressed concrete box-girders, and prestressed concrete beams, and to demonstrate the procedure,
reinforced concrete channel-girders. it is used to predict the shear strength of concrete girders in three
existing bridges. The current paper deals only with members that
Keywords: bridge(s); girders; shear strength. contain at least minimum transverse reinforcement (stirrups),
whereas a previous paper6 deals with members that have less
INTRODUCTION than minimum stirrups.
The strength of concrete bridge girders needs to be evaluated
to determine the load capacity rating of bridges because of BACKGROUND
increased traffic loads or deterioration of bridges. The shear The shear resistance of a structural concrete member has
strength of concrete bridge girders commonly limits the load traditionally been expressed as the sum a concrete contribution
capacity ratings of bridges. Vc, a stirrup contribution Vs, and a contribution from inclined
The shear strength of structural concrete is a complex prestressing Vp. The nominal shear resistance can be
phenomenon. Building codes include simplifications that expressed in a generalized form as follows
generally result in safe designs. The additional construction
costs are justified by the reduced chance of a design error. A v f y d v cot θ
V = V c + V s + V p = β f c′ b w d v + --------------------------
- + Vp (1)
On the other hand, the consequence of these same simplifi- s
cations may be greater when a simplified shear design
method is used to evaluate existing girders that cannot be where β is the concrete contribution factor, fc′ is the specified
made a little stronger. The simplifications may result in compressive strength of concrete, bw is the web width, dv is
unnecessary load restrictions on bridges or unnecessary effective shear depth of the member, Av is the cross-sectional
repairs of bridge girders. Thus, more complex procedures are area of stirrups within a distance s along member, fy is the
justified for shear strength evaluation. yield strength of stirrups (which are assumed to be
According to shear design procedures based on the modified yielding), and θ is the inclination of concrete principal
compression field theory (MCFT),1 such as AASHTO compressive stress (measured from the longitudinal axis).
LRFD2,3 and the 2006 Canadian Highway Bridge Design In North America, the concrete contribution has traditionally
Code (CHBDC),4 the shear strength of a girder is a function been taken equal to the shear strength of a member without
of axial strain, which depends on a number of factors stirrups, which is equal to the shear force at first diagonal
including the applied shear force. Using such a design procedure cracking.7 For prestressed concrete members, β depends on
to evaluate strength requires trial and error, as the applied the moment-to-shear ratio and level of prestress. For a
shear force at failure is needed to calculate shear strength. reinforced concrete member with no size effect (small size
An alternative approach for shear strength evaluation is to or with minimum stirrups), a safe value for Vc results from
use a computer program such as Response-2000,5 which assuming β = 2 psi (0.17 MPa) and dv = d. The stirrup
applies the MCFT using a fiber model approach. This would
provide considerable insight such as whether the failure
mode will be ductile (due to reinforcement yielding), or ACI Structural Journal, V. 106, No. 4, July-August 2009.
MS No. S-2007-208.R2 received July 5, 2008, and reviewed under Institute publication
more brittle (due to diagonal crushing of concrete). While policies. Copyright © 2009, American Concrete Institute. All rights reserved, including the
such computer methods are very useful for special investigations, making of copies unless permission is obtained from the copyright proprietors. Pertinent
discussion including author’s closure, if any, will be published in the May-June 2010 ACI
the complexity of data input and output makes it difficult to use Structural Journal if the discussion is received by January 1, 2010.

416 ACI Structural Journal/July-August 2009

ACI member Afshin Esfandiari is a Senior Structural Engineer at WorleyParsons
Westmar, North Vancouver, BC, Canada. He received his BS in civil engineering from
Bahai Institute for Higher Education (BIHE), Tehran, Iran; his MASc in structural
engineering from Carleton University, Ottawa, ON, Canada; and his PhD from the
University of British Columbia, Vancouver, BC, Canada.

Perry Adebar, FACI, is a Professor of Structural Engineering at the University of British

Columbia. He is a member of ACI Committees 341, Earthquake-Resistant Concrete
Bridges; 374, Performance-Based Seismic Design of Concrete Buildings; and E803,
Faculty Network Coordinating Committee; Joint ACI-ASCE Committees 441,
Reinforced Concrete Columns, and 445, Shear and Torsion; and is a member of the
Canadian Code Committee CSA A23.3.

contribution Vs that is added to the aforementioned Vc is

calculated assuming θ = 45 degrees (cotθ = 1.0), and dv = d.

UNIFORM SHEAR ELEMENTS

The simplest shear problem involves an element with
uniformly spaced longitudinal (x-direction) and transverse Fig. 1—MCFT predictions of shear response for different
(z-direction) reinforcement, no inclined prestressing tendons, uniform shear elements.
and subjected to uniform applied normal stresses fx and fz
and shear stress v. Such an element will have uniform 2007 AASHTO LRFD
average reinforcement stresses fsx and fsz , and uniform The shear design method given in the 2007 AASHTO
average concrete stresses fcx, fcz , and vc = v. Equation (1) can LRFD2 can be used to calculate the shear strength of a
be modified for such an element by eliminating Vp and uniform shear element. Derivation of this method from the
dividing the remaining shear force components by the shear MCFT was presented by Collins et al.8 The method includes
area bwdv , resulting in tables with β and θ as functions of longitudinal strain εx and
shear stress ratio v/fc′. For εx from –0.0002 to +0.001, and v/fc′
v = v c + v s = β f c′ + ρ z fycot θ (2) from 0.075 to 0.250, θ varies from 22 to 37 degrees and β
varies from 1.5 to 6.3 psi (0.13 to 0.53 MPa).
The desired shear strength of an element can be achieved
where ρz is the transverse reinforcement ratio Av/bw s. with different reinforcement amounts. For example, an
The MCFT1 was developed from testing reinforced element with less ρz and more ρx will have a lower Vs and
concrete elements subjected to uniform shear stress. It is a higher Vc, but may have the same total shear strength. The
smeared, rotating crack model where the inclination of diagonal values of θ and β provided in the tables for members with at
cracks is determined by combining equilibrium requirements, least minimum transverse reinforcement correspond to one
strain compatibility assumptions, and empirical average particular solution.9 Other tables can be developed for
stress/average strain relationships for cracked concrete and different relative amounts of reinforcement.
reinforcement. The MCFT can be used to determine β and θ
in Eq. (1) and (2) by assuming no z-direction normal stress
2006 CHBDC and 2008 AASHTO LRFD
is applied on the element (fz = 0). According to MCFT, the
The shear design methods in the 2006 CHBDC4 and 2008
concrete contribution Vc is the shear in addition to that
interim of AASHTO LRFD3 are also based on MCFT, but
resisted by yielding transverse reinforcement that can be
include simplified equations rather than the tables for θ and
transferred across diagonal cracks by rough crack surfaces.
β. For members with at least minimum transverse reinforcement,
The MCFT can be used to predict the shear stress/shear θ and β depend only on the longitudinal strain εx . The
strain relationships of structural concrete elements with following expressions are given in the 2006 CHBDC4
different amounts of transverse and longitudinal reinforcement,
as shown in Fig. 1. The applied shear stress and resulting
β = -------------------------- psi ⎛ β = --------------------------MPa⎞
shear strain are in the directions parallel to the orthogonal 4.8 0.4
(3)
reinforcement. Until concrete cracks, the inclination of diagonal 1 + 1500ε x ⎝ 1 + 1500ε x ⎠
compression θ in elements subjected to pure shear is 45 degrees.
After cracking, θ reduces depending on the relative amounts
θ = 29 + 7000εx (4)
of reinforcement ρz /ρx and the shear stress ratio v/fc′. After
the transverse (z-direction) reinforcement yields, the shear
strain γ increases, the longitudinal strain εx increases, the Over the possible range of εx = 0 to 0.001 (with 58 ksi
inclination of diagonal compression θ reduces, and the [400 MPa] grade longitudinal reinforcement), θ varies from
concrete contribution factor β reduces. As vc reduces and vs 29 to 36 degrees and β varies from 1.9 to 4.8 psi (0.16 to 0.4 MPa).
increases, the total shear stress may reduce, stay constant, or The equations give one particular solution that is independent
increase, depending on the amount of transverse reinforcement, of shear stress ratio v/fc′ because β was developed for the
as shown in Fig. 1. The point at which diagonal compression case of no transverse reinforcement (low shear stress),
stresses in concrete reaches the crushing strength of concrete and θ was developed for the maximum shear stress ratio
is also shown on each curve. In the element with ρz = 0.005 v/fc′ = 0.25. Combining θ and β equations results in a simpler
and ρx = 0.02, all reinforcement yields at a shear strain of design procedure that is usually safe.10,11 The reference strain
approximately 0.008. εx used in the 2008 AASHTO LRFD3 is exactly twice the

ACI Structural Journal/July-August 2009 417

 calculated from the applied forces and actual amount of
longitudinal reinforcement and prestressing. The procedure
also includes a check to ensure the required longitudinal
reinforcement is available to equilibrate the longitudinal
compression force in concrete.
Three main shear failure modes are: 1) yielding of transverse
reinforcement; 2) crushing of concrete after yielding of
transverse reinforcement; and 3) yielding of longitudinal
reinforcement. A fourth mode that only controls the shear
strength of girders with very large percentages of transverse
reinforcement is concrete crushing before reinforcement
yields, and this mode can be avoided by limiting the shear
stress from Eq. (2) to 0.25fc′. The three main shear failure
modes are illustrated in Fig. 1. The element with ρz = 0.002
and ρx = 0.03 reaches maximum shear immediately upon
yielding of transverse reinforcement, whereas the two other
elements with ρx = 0.03 gain strength after transverse
reinforcement yields and reach a maximum shear at concrete
crushing. For a particular element, these two modes must be
evaluated and the larger controls the strength. The element
with ρz = 0.005 and ρx = 0.02 reaches maximum shear when
longitudinal reinforcement yields. The shear strength of an
element is limited to the shear associated with this third mode.
Figure 2 shows the relationship between θ and εx for
different transverse reinforcement amounts ρz. Solid lines
show the relationships given by the MCFT, which is
approximately linear for a constant ρz. The relationships are
very different at first yielding of transverse reinforcement
(Fig. 2(a)) and concrete crushing after yielding (Fig. 2(b)).
The 2006 CHBDC4 method has one equation for θ as a
function of εx , shown in Fig. 2 as a dotted line. As the 2006
CHBDC4 approximate value for θ is generally larger than
the MCFT value, it generally predicts a smaller Vs than the
MCFT. This is not the case when εx is large at yielding of
transverse reinforcement (right-hand side of Fig. 2(a)), but
there is a significant strength increase after transverse
reinforcement yielding for large εx.12
As indicated in the 2007 AASHTO2 method, θ depends on
the principal compression strain, which is related to the shear
stress ratio v/fc′. This ratio is known in design, but not during
evaluation until the shear strength is determined. Thus ρz fy/fc′
was used to develop the following expressions that fit the
trend predicted by the MCFT
Fig. 2—Comparison of predicted angle θ with MCFT results
at: (a) first yielding of transverse reinforcement; and (b) θ = θo + Δθεx (5)
concrete crushing after yielding. fc′ = 5.8 ksi (40 MPa), fy =
58 ksi (400 MPa). where at yielding of transverse reinforcement

value used in the 2006 CHBDC,4 and thus 1500εx in Eq. (3) ρz fy
θ o = ⎛ 85 --------
- + 19.3⎞ ( – 50ε y + 1.1 ) (6)
is replaced by 750εx , and 7000εx in Eq. (4) is replaced by ⎝ f c′ ⎠
3500εx.The 2008 AASHTO LRFD3 also has a simple design
procedure that is not meant for strength evaluation and thus is
not discussed. Δθ = 1000[37.5(–200εy + 1.4) – θo] (7)

Proposed expressions for strength evaluation and at concrete crushing

As was done8,10 to develop all shear design methods
based on the MCFT, the diagonal crack spacing is assumed
ρz fy
to be 12 in. (300 mm) in members with at least minimum θ o = 119 --------
- + 15.6 (8)
transverse reinforcement. To develop equations that later can f c′
be applied to beams, a value is assumed for the uniform
longitudinal strain εx of the element, and the corresponding ρz fy
concrete longitudinal compression stress is calculated. When Δθ = 15,000 --------
- + 2000 (9)
applying the procedure, the actual longitudinal strain εx is f c′

418 ACI Structural Journal/July-August 2009

Fig. 3—Comparison of predicted concrete contribution with Fig. 4—Comparison of predicted shear stress with MCFT
MCFT at: (a) first yielding of transverse reinforcement; and at: (a) first yielding of transverse reinforcement; and (b)
(b) concrete crushing after yielding. fc′ = 5.8 ksi (40 MPa), concrete crushing after yielding. fc′ = 5.8 ksi (40 MPa), fy =
fy = 58 ksi (400 MPa). 58 ksi (400 MPa).

The angles predicted by these equations are also shown in is used in the proposed method for 58 ksi (400 MPa) and
Fig. 2 as dashed lines, and there is good agreement with higher-grade steel. For lower-grade transverse reinforcement, a
MCFT, which validates the approach of using ρz fy/fc′. The larger concrete contribution at yielding of transverse
additional validation for other reinforcement and concrete reinforcement is appropriate due to reduced strains, and the
strengths can be found in Reference 12. constant β given previously should be multiplied by (–300εy +
When the longitudinal strain equals the transverse strain, 1.6), where εy is the yield strain of the transverse reinforcement.
the MCFT angle must equal 45 degrees. For 58 ksi (400 MPa) As β varies more at concrete crushing after yielding, it is
grade reinforcement, the transverse reinforcement yields at a proposed that at concrete crushing
strain of 0.002, thus the angles converge to 45 degrees in
Fig. 2(a) at a longitudinal strain of 0.002. The largest ρz fy ρz fy
- + 0.36 psi ⎛ β = 0.65 --------
β = 7.8 -------- - MPa⎞ (10)
longitudinal strain used in the shear analysis of beams with f c′ ⎝ f c′ ⎠
58 ksi (400 MPa) grade longitudinal reinforcement is 0.001,
and at this strain, the MCFT angles have almost converged.
For simplicity, Eq. (6) and (7) have the angles converging at Figure 3(b) shows that the proposed expression agrees well
a longitudinal strain of 0.001, and for 58 ksi (400 MPa) grade with the MCFT and is safe.
reinforcement, that angle is 37.5 degrees. Total shear stress v is the sum of vc , which depends on β,
The solid lines in Fig. 3 show the concrete contribution and vs, which depends on θ. Figure 4 compares the MCFT
factor β according to the MCFT at first yielding of transverse total shear stress with results of the proposed method at
reinforcement (Fig. 3(a)) and concrete crushing after transverse reinforcement yielding (Fig. 4(a)) and concrete
yielding of transverse reinforcement (Fig. 3(b)). The 2006 crushing after transverse reinforcement yielding (Fig. 4(b)).
CHBDC4 approximate β value is shown as a dotted line. As The resultant of longitudinal concrete compression stress
this equation was developed for members without transverse fcx is the axial compression force Nv required in a beam web
reinforcement, it gives a larger β than the MCFT for low εx to resist shear, thus this stress is also referred to as nv. In the
values in members with minimum transverse reinforcement. 2007 AASHTO LRFD,2 it is assumed that nv = vcotθ,
The larger estimate of Vc in these members compensates for whereas in the 2006 CHBDC4 and the 2008 AASHTO-LRFD,3
the smaller estimate of Vs.11 this is further simplified to nv = 2v. According to the MCFT,
the relationship is
Figure 3(a) indicates the MCFT concrete contribution
factor β does not vary significantly at transverse reinforcement
yielding, and thus a constant value of β = 2.17 psi (0.18 MPa) nv = 2vccot2θ + ρz fycot2θ (11)

ACI Structural Journal/July-August 2009 419

 Vc factor β for those cases, as shown in Fig. 3(a). At concrete
crushing, the appropriate simplification12 is to use Eq. (13) and
(14) if θo ≤ 23 degrees and Eq. (15) and (16) if θo > 23 degrees.
The nv calculated from the MCFT, as well as calculated
from Eq. (12) to (16) using the aforementioned simplifications, is
compared at yielding of transverse reinforcement in Fig. 5(a)
and concrete crushing in Fig. 5(b). The 2006 CHBDC4
approximation of nv = 2v is generally safe (larger estimate of
nv than actual) at yielding of transverse reinforcement. Thus,
the estimated axial tension force in the longitudinal reinforcement
that balances the axial compression force Nv will be larger
than actual, and the estimated axial strain εx will be larger
than actual. When this trend is combined with the trend
shown in Fig. 4, the 2006 CHBDC4 method generally gives
safe predictions of shear strength for a member with a given
amount of longitudinal reinforcement.
There are cases when an element cannot reach the shear at
first yielding of transverse reinforcement or concrete
crushing after yielding because the longitudinal (x-direction)
reinforcement does not have the capacity to equilibrate the
longitudinal compression force in concrete. The average
longitudinal compression stress nv, which influences the strain,
was previously discussed. This average compression stress is
smaller due to concrete resisting some average tension stresses
between cracks. The axial compression stress nv* required
locally at diagonal cracks to resist shear is larger

nv* = 2vccotθ + ρz fycot2θ (17)

Fig. 5—Comparison of predicted longitudinal compression
stress with MCFT at: (a) first yielding of transverse The values of vc and θ determined at transverse reinforcement
reinforcement; and (b) concrete crushing after yielding. fc′ yielding and concrete crushing can be used in Eq. (17) to
= 5.8 ksi (40 MPa), fy = 58 ksi (400 MPa). determine the longitudinal compression that must be
balanced by tension in longitudinal reinforcement at
diagonal cracks. A safe estimate of shear to cause both
To simplify this equation in the proposed method, cot2θ
longitudinal and transverse reinforcement to yield results
and cot2θ have been approximated as linear functions12 for
from assuming the concrete contribution is negligible.
θ > 25 degrees and θ ≤ 25 degrees. Substituting these linear
Setting vc = 0 in Eq. (2) and (17), and solving these two equations
functions, as well as vc = β f c′ , and Eq. (5) for θ as a
for v (eliminating θ), results in the following expression for shear
function of εx into Eq. (11) gives
strength given the maximum nv* controlled by the capacity of
distributed longitudinal reinforcement denoted nvc
nv = nvo + Δnvεx (12)

where for θ > 25 degrees v = ρ z f y n vc (18)

Δnv = (–0.09β f c′ – 0.20ρz fy)Δθ (13) If 10% of the shear strength is actually coming from a
concrete contribution, solving for the shear strength from Eq. (2)
and (17) results in a 0.5% increase over that estimated from
Δn Eq. (18). If vc is 20% and 30% of v, the difference increases
n vo = ---------v θ o + 40β f c′ + 9.4ρ z f y (14) to 2% and 5%, respectively. Thus, Eq. (18) is a reasonable
Δθ
approximation for most cases with transverse reinforcement.
and for θ ≤ 25 degrees
APPLICATION TO BEAMS
Response-2000
Δnv = (–0.15β f c′ – 0.77ρz fy)Δθ (15) Response-20005 applies the MCFT to beams using a
multi-layer analysis. In their report on shear strength of
bridge girders, Hawkins et al.13 compared Response-2000
Δn
n vo = ---------v θ o + 5.5β f c′ + 23.6ρ z f y (16) predictions with the results of 149 tests (including 85
Δθ prestressed concrete girders) that contained minimum stirrups
(ρz fy > 50 psi [0.35 MPa]). They found mean ratios of
A simplification that avoids trial-and-error is to use only measured-to-predicted shear strength of 1.02 and 1.11, and
Eq. (13) and (14) at transverse reinforcement yielding. coefficients of variation (COV) of 11% and 17% for reinforced
Although this results in low estimates of nv when θ is less and prestress girders, respectively.
than 25 degrees, which results in smaller εx and θ and, thus, Figure 6 shows the Response-2000 predictions for a
a larger Vs, this is compensated for by the low estimate of the typical prestressed I-girder at the point of stirrups first

420 ACI Structural Journal/July-August 2009

 Fig. 6—Variation of shear response over depth of prestressed I-girder with composite
deck slab: (a) cross section; (b) longitudinal strain; (c) shear flow; (d) principal
compression stress angle θ; and (e) longitudinal normal stress multiplied by beam
width. Refer to Fig. 9(a) for location of section.

yielding and concrete crushing after yielding. The longitudinal assumed to be uniform over the shear depth dv, which is
strains vary linearly over the depth, as shown in Fig. 6(b). estimated as 0.9d. In all methods except the 2008
The transverse shear flow (Fig. 6(c)) varies in a complex AASHTO-LRFD,3 the longitudinal strain εx used in the
nonlinear way with the maximum value being in the deck shear analysis is taken as the average value over the
slab, which is composite with the girder. The inclination of section, and for simplicity, this is estimated as half the
principal compression stress (Fig. 6(d)) varies from 0 degrees on strain of the flexural tension reinforcement assuming the
the top of the deck slab to 90 degrees at the bottom face of concrete surrounding this reinforcement is cracked
the girder. Over the height of the web, the angle generally
varies between 27 and 39 degrees at stirrup yielding and M/jd + 0.5N v – f p A p
-≥0
ε x = ------------------------------------------------ (19)
between 24 and 26 degrees at concrete crushing. The longitudinal 2 ( Es As + Ep Ap )
concrete normal stress is multiplied by member width (analogous
to shear flow) in Fig. 6(e) to facilitate comparison with the
where M is the bending moment at the section of interest; jd
prediction, which assumes a constant web width.
is the internal flexural lever arm (M/jd is the flexural tension
force); Nv is the axial compression force needed to resist
Simplified procedures for shear design shear in the web taken as nvbwdv; fp Ap is the effective
The shear design procedures in AASHTO LRFD3 and prestressing force; As and Ap are the area of nonprestressed
2006 CHBDC4 are based on a single shear analysis over the and prestressed flexural tension reinforcement, respectively;
section. The shear stress v, inclination of principal compression and Es and Ep are the modulus of elasticity of nonprestressed
stress θ, and longitudinal compression stress nv are all and prestressed reinforcement, respectively. Equation (19) is

ACI Structural Journal/July-August 2009 421

valid only if the resulting strain (net force in numerator) is 2
where ς = 2λ ε x E p A pw + λf p A pw + α f c′ A tf + f p A p , and
not less than 0. When the strain is compressive, the concrete
compression stiffness must be added to the denominator of rearranging this equation to solve for εx gives
Eq. (19). Alternatively, εx can safely be taken as 0. In the
2008 AASHTO-LRFD,3 the longitudinal strain εx used in M/jd + 0.5n vo b w d nv – α f c′ A tf – f p ( A p + λA pw )
the shear analysis is taken as the strain of the flexural tension ε x = ---------------------------------------------------------------------------------------------------------------------------------
2
- (21)
2 [ E s ( A s + 0.25A sw ) + E p ( A p + λ A pw ) ] – 0.5Δn v b w d nv
reinforcement (twice εx given by Eq. (19)).

Proposed procedure for shear strength evaluation Note that the area of flexural tension reinforcement As in
The proposed method also uses a single shear analysis the denominator of Eq. (21) is multiplied by a factor of 4
over the section depth and the assumption of uniform shear compared to the area of longitudinal reinforcement Asw
stress over the shear depth dv. The actual shear stress is not centered in the web. A factor of 2 comes from the assumption
uniform, as shown in Fig. 6(c); however, the shear stress at that midheight strain is half the strain of the flexural tension
midheight is a reasonable estimate of average shear stress. In reinforcement and a second factor of 2 comes from the need
the simplified design procedures, the longitudinal concrete to provide twice as much reinforcement at section mid-depth
compression stress nv required to resist shear is also assumed to resist bending compared to reinforcement on the flexural
to be uniform over the shear depth dv. Figure 6(e) indicates tension face.
that nv at midheight is a reasonable estimate of nv over the The assumption that longitudinal strain at section
web region of the member, but is not a good estimate of the midheight is half the strain of the flexural tension reinforcement is
average nv over the complete shear depth. The shear depth generally safe, as the strain on the opposite face is usually
extends well into the deck slab, and as described previously, compressive (refer to Fig. 6(b)). If the bending moment is
the maximum shear flow occurs in this region; however, this small and the shear force is large, the section may be
portion of the member does not experience any diagonal subjected to tension strains over the full depth. This would be
cracking and thus does not develop the additional longitudinal the case if
concrete compressive stresses nv due to shear. The flexural
compression in the deck slab that is balanced by flexural tension M/jd ≤ 0.5nvobwdnv + εx[0.5Δnvbwdnv – (22)
should not be part of Nv. See, for example, the compression
stress cutoff at –12 kip/in. (–2068 N/mm) at the top of Fig. 6(e). 0.5EsAsw – 2λ(1 – λ)EpApw] – (1 – λ)fpApw
The concrete longitudinal compressive stresses due to
shear nv also do not extend down into the flexural tension
In that case, the section midheight strain calculated by
flange where concrete is in tension (refer to the bottom of
Eq. (21) should be multiplied by 2.
Fig. 6(e)). Thus, in the proposed method, nv estimated at
section midheight is assumed to be uniform over a reduced Yielding of longitudinal reinforcement limits the shear
depth dnv from the top of the bottom flange to the bottom of strength to
the top flange (refer to Fig. 6(a) and (e)). The tension
stiffening effect of concrete in the tension flange, which has V ≤ ρ z f y b w d v N vc + Vp (23)
area Atf of 6 x 24 in. (150 x 610 mm) in Fig. 6(a) can be
accounted for by assuming an average concrete tension stress of where
α f c′ over this area, where α = 1.8 psi (0.15 MPa).12
Another refinement in the proposed method is to rigorously Nvc ≤ 2[fy(As + 0.5Asw) + fpr(Ap + λApw) – M/jd] (24)
account for reinforcement in the web. The tension force
resisted by nonprestressed reinforcement uniformly distributed and fpr is the maximum stress in the prestressing tendons.
over the web is εxEs Asw, where Asw is the total area of Equation (24) must be evaluated separately for the flexural
reinforcement in the web, and the average strain of this tension and flexural compression sides of the member, and
reinforcement is equal to the section midheight strain εx. the smaller value controls the shear strength. The variables
Prestressing tendons in the web may be located at any elevation. As and Ap are the areas of longitudinal reinforcement on the
The distance from the flexural compression face to the flexural tension or flexural compression side of the member,
centroid of flexural tension reinforcement is denoted d, while Asw is centered in the web. The location of Apw is
while the effective depth to the centroid of the prestressing accounted for by substituting (1 – λ) for λ when applying
tendons is denoted dpw. The parameter is used to account for the Eq. (24) to the flexural compression side. The force M/jd
location of tendons relative to the flexural tension reinforcement. in Eq. (24) shall be taken as positive on the flexural tension
For draped tendons, it is safe to ignore inclination of tendons in side and negative on the flexural compression side. The
calculation of both longitudinal strain and horizontal force compression force Nv required to resist shear in a diagonally
component12; however, the vertical component of the tendon cracked web is resisted by distributed reinforcement in the
force is added to shear strength as in Eq. (1). web Asw, Apw (if present), and the remaining capacity of
A revised Eq. (19) results from using Eq. (12) to estimate nv flexural tension reinforcement (As , Ap).
at section midheight, using the reduced depth dnv to calculate The internal flexural lever-arm jd can be taken as 0.9d
Nv, accounting for the tension force resisted by distributed when estimating the strain from Eq. (21); however, when
reinforcement centered in the web, Asw, accounting for the loca- checking yielding of longitudinal reinforcement in Eq. (23)
tion of prestressing tendons in the web; and the force resisted by and (24), a more accurate estimate is needed
average concrete tension stresses in the tension flange
f pr A p + f y A s
M/jd + 0.5 ( n vo + Δn v ε x )b w d nv – 0.5ε x E s A sw – ς jd = d – ---------------------------
- (25)
ε x = -----------------------------------------------------------------------------------------------------------------
- (20) 1.2f c′ b f
2 ( Es As + Ep Ap )

422 ACI Structural Journal/July-August 2009

where bf is the width of the flexural compression zone and
0.6fc′ is an appropriate average flexural compression stress at
first yielding of longitudinal reinforcement.12

VERIFICATION: COMPARISON WITH TEST

RESULTS
To verify the proposed evaluation method, shear strength
predictions for 80 reinforced concrete beams and 88 prestressed
concrete beams with at least minimum stirrups were
compared with measured strengths. The tests include ten
73 in. (1854 mm) deep prestressed I-girders14 and seven
44 in. (1118 mm) deep prestressed I-girders.15 The
remaining 151 beams, which had a minimum depth of 12 in.
(300 mm), were selected from a shear database,14 and
literature.16-18 A summary of all specimens used for the
verification is given in Reference 12.
The shear strengths of the beams were predicted by the
proposed evaluation procedure as well as the shear design
provisions from the 2007 and 2008 AASHTO-LRFD2,3 and
ACI 318-05.19 As mentioned previously, the 2006 CHBDC4
gives identical results to the 2008 AASHTO-LRFD.3 The
loading for most tests consisted of concentrated loads, and
the critical shear sections for these were taken at d from the
concentrated loads toward the supports. For tests with
uniformly distributed loading, the critical sections were
taken at d from the supports or locations of stirrup spacing
change toward midspan. The appropriate Vp was added to the
shear strength of beams with inclined tendons. A complete
summary of these calculations is given elsewhere.12
Figures 7(a) and (b) present the cumulative frequency
distributions of test-to-predicted shear strength ratios of 80
reinforced concrete (RC) beams and 88 prestressed concrete
(PC) beams, respectively. These diagrams show the number
of ratios (as a portion of the total number) that are equal to or
Fig. 7—Cumulative frequency of test-to-predicted ratios of
less than the corresponding test-to-predicted ratio on the
proposed and code methods: (a) reinforced concrete beams;
horizontal axis. A “perfect prediction” has a cumulative
and (b) prestressed concrete beams.
frequency of zero for a test-to-predicted ratio less than 1.0,
and a cumulative frequency of 1.0 for a test-to-predicted
ratio greater than 1.0. Figure 7(b) summarizes the predictions for 88 prestressed
The cumulative frequency plots provide much information, concrete beams, and a subset of 22 prestressed concrete
but only a few points will be discussed. Looking at predictions beams with εx greater than 0.0001. While the ACI 318-0519
for all 80 reinforced concrete beams (Fig. 7(a)), the proposed shear design method tends to be overly safe for reinforced
method has the largest (safest) minimum test-to-predicted concrete beams, it is very unsafe for large prestressed
shear strength ratio of 0.87, whereas the 2007 AASHTO- concrete beams where Vcw controls the shear strength, as
LRFD2 has the smallest value—0.76. All three methods based previously recognized.13 Approximately 50% of ACI 318-0519
on the MCFT have approximately 20% of predictions predictions for the 88 beams have test-to-predicted ratios
resulting in test-to-predicted ratios less than 1.0; however, for less than 1.0. The improvement of the predictions from the
the proposed method, half of these are greater than 0.95. The proposed method compared to the two MCFT-based shear
median test-to-predicted ratio is 1.12 for the proposed method design procedures is evident for prestressed concrete beams.
and the 2007 AASHTO-LRFD,2 1.16 for the 2008 AASHTO- For example, the median test-to-predicted ratio of 88 beams
LRFD,3 and 1.28 for ACI 318-05.19 Prediction variability, is 1.25 from the proposed method versus 1.31 and 1.33 for
which influences the steepness of cumulative frequency plots, the 2008 and 2007 AASHTO-LRFD,3,2 respectively.
is commonly expressed using the COV. The proposed method
has the smallest COV of test-to-predicted ratios equal to EXAMPLES
16.5%, whereas ACI 318-0519 has the largest at 18.7%. The proposed procedure was used to evaluate the shear
It is interesting to examine how predictions change when strength of three existing concrete bridges. The first was a
only the 26 beams with at least twice the minimum stirrups 68.9 ft (21 m) single-span bridge with six prestressed
are considered. For all but the proposed method, there is a concrete I-girders (Fig. 8(a)) spaced at 78.7 in. (2 m). The
significant increase in the test-to-predicted ratios (lines shift second bridge consisted of nine prestressed box girders
right). The median of test-to-predicted ratios for the (Fig. 8(b)) spanning 97.8 ft (29.8 m). The 20 prestressing
proposed method is essentially unchanged at 1.13 and the tendons in the web were draped. The final bridge
COV reduces to 9.1%, whereas the median ratios for the consisted of 14 precast nonprestressed channel girders
2008 AASHTO-LRFD,3 the 2007 AASHTO-LRFD,2 and (Fig. 8(c)) interconnected by grouted reinforcing bars. Each
ACI 318-0519 are 1.18, 1.24, and 1.42, respectively. simple span of this multispan bridge was 27.6 ft (8.40 m).

ACI Structural Journal/July-August 2009 423

Fig. 8—Cross sections of concrete girders from bridge
examples.

The stirrup spacings were constant in the box girder and Fig. 9—Comparison of predicted shear strengths along half-
are shown in Fig. 8(b). The stirrup spacings varied in the span of simply supported I-girder bridge: (a) shear force
other two girders, and can be calculated using the area per envelope, bending moment envelope, and quantity of stirrups;
stirrup in Fig. 8 and reinforcement ratios shown in Fig. 9(a) (b) Response-2000 and proposed method predictions; and (c)
and Table 1. The following material properties were used to Response-2000 and code design method predictions.
evaluate all three girders: fc′ > = 5.8 ksi (40 MPa); fy = 58 ksi
(400 MPa); fpu = 270 ksi (1860 MPa); and Ep = Es = 29,000 ksi
(200,000 MPa). The variation of predicted strengths over half the span of
To compare different codes, nominal resistances were the I-girder is shown in Fig. 9. At locations where spacing of
calculated. The highway truck loading (three different trucks stirrups changed, the quantity Av /s was assumed to vary
were used12) and distribution of loading to individual girders linearly over the length d centered on the location where
were done according to the 2006 CHBDC.4 The trucks were spacing changed. This occurred at 24 and 28.5 ft (7.32 and
moved separately along the bridge to determine live load 8.69 m) from the support. Figure 9(b) compares the shear
shear forces and bending moments that were then distributed strength determined from Response-2000 and the proposed
transversely among girders based on the CHBDC4 procedure, method for concrete crushing mode (solid lines) and stirrup
which depends on such factors as lane width and girder yielding mode (dashed lines). Generally, there is very good
spacing. Live load shear force and bending moments were agreement along the span. Note the information presented in
combined with dead load effects using CHBDC4 load factors. Fig. 6 is for the I-girder at 26 ft (7.92 m) from the support.
The resulting shear force and bending moment envelopes are Figure 9(c) compares the shear strength from Response-2000
shown in Fig. 9(a) for the I-girders, and the values at three with shear strengths according to the 2007 AASHTO-LRFD,2
sections of each bridge are given in Table 1. the 2006 CHBDC4 (same as the 2008 AASHTO-LRFD3),

424 ACI Structural Journal/July-August 2009

Table 1—Summary of predicted shear strengths for bridge examples
Predicted shear strength, kip
Section information (Response-2000 shear-to-predicted-shear ratio) Response-2000 shear
Distance Proposed strength, kip
from AASHTO CHBDC
Bridge support, ft ρz , % Vu, kip Mu, kip·ft 2007 2006 ACI 318 Yielding* Crushing† Yielding* Crushing†
4.9 0.87 176.2 894 259.8 (1.01) 283.7 (0.92) 276.7 (0.95) 260.5‡(1.01) 248.4 261.9 262.2‡
I-girder 17.2 0.87 124.1 2370 186.7 (1.22) 174.6 (1.30) 249.1 (0.91) 194.5 215.7‡ (1.05) 197.4 227.3‡
‡
26.0 0.44 93.3 2740 105.7(1.40) 99.2 (1.49) 141.2 (1.05) 110.4 133.0 (1.11) 120.4 148.1‡
19.5 0.80 115.7 1188 380.6 (1.04) 406.4 (0.98) 409.3 (0.97) 379.9‡ (1.05) 363.0 397.2‡ 388.3
‡
Box girder 29.3 0.80 92.2 2086 313.1 (1.12) 303.2 (1.16) 295.9 (1.19) 317.7 334.5 (1.05) 346.8 351.1‡
39.1 0.80 68.4 2660 278.3 (1.17) 259.9 (1.25) 254.9 (1.27) 279.0 314.4‡ (1.03) 306.6 324.3‡
‡
2.1 0.59 35.5 96 110.6 (1.22) 103.9 (1.30) 104.7 (1.30) 130.1 135.6 (1.00) 129.6 135.0‡
Channel girder 4.7 0.29 31.3 153 84.4 (1.14) 78.9 (1.22) 68.8 (1.40) 81.0 91.7‡ (1.05) 77.5 96.5‡
11.0 0.10 18.8 242 47.0 (0.95) 43.4 (1.03) 44.3 (0.99) ‡ 37.2 ‡ 43.0
46.3 (0.96) 44.5
Response-2000 shear strength-to-predicted Mean 1.14 1.18 1.11 1.04 —
strength ratio COV, % 11.8 15.4 15.9 4.4 —
*
Yielding = first yielding of stirrups.
†
Crushing = crushing of concrete after yielding of stirrups.
‡
Governing failure mode.
Note: COV = coefficient of variation; 1 ft = 0.3048 m; 1 kip = 4.448 kN.

and ACI 318-05.19 The 2007 AASHTO-LFRD3 gives a safe procedure can be implemented into a small computer
prediction all along the span. The 2006 CHBDC4/2008 program for checking numerous sections along a bridge
AASHTO-LRFD3 shear design method gives a high prediction girder and the results can easily be verified by hand calculation.
near the support where axial strains are very low. On the A previous paper6 presents a similar procedure for members
other hand, the method gives a low prediction at 13.8 ft that have less than minimum stirrups.
(4.21 m) from the support where the predicted strength is The shear design methods in AASHTO-LRFD2,3 and the
175 kip (778 kN). The shear strength according to 2006 CHBDC,4 which are based on the MCFT, require trial-
Response-2000 is 211 kip (938 kN) at stirrup yielding and-error for strength evaluation. The proposed evaluation
and 232 kip (1032 kN) at concrete crushing. Near midspan, the procedure was developed so that trial-and-error is not
2006 CHBDC4/2008 AASHTO-LFRD3 prediction corresponds required, but also includes a number of refinements such as
well with stirrup yielding. The ACI 318-0519 shear design accounting for: 1) the influence of Vc (concrete tension
method again gives very unsafe predictions where Vcw controls stresses) on the average longitudinal compression force Nv
shear strength for the first 20 ft (6.10 m) from the support. required to resist shear in a diagonally cracked web (Vc
The predicted shear strengths at three sections along each reduces average tension strain of a member); 2) the difference
of the three bridge girders are summarized in Table 1. For between total shear depth dv and depth of diagonally cracked
each bridge, one section was located in the low-moment web dnv; 3) the tension force resisted by distributed longitudinal
region close to the support, another section was located in the reinforcement in web; 4) location of prestressed tendons in
high-moment region near midspan, and the third section was web; and 5) tension-stiffening by cracked concrete around
located between the other two. The shear strength at stirrup flexural reinforcement.
yielding and concrete crushing according to Response-2000 To validate the proposed procedure, shear strength
are shown, and the critical one is identified thus (‡) in Table 1. predictions were compared with results from tests on 168 beams.
Unlike the design methods, the proposed procedure also Predicted shear strengths from the proposed procedure were found
gives two shear strengths, and these generally agree well to be in better agreement with test results than predicted strengths
with the Response-2000 results. The ratios of Response-2000 from AASHTO-LRFD,2,3 CHBDC,4 and ACI 318-05.19
shear-strengths-to-predicted-shear-strengths are shown in To demonstrate the evaluation procedure, it was used to
brackets. The nine governing ratios from the proposed evaluate shear strengths of three different types of concrete
method vary from 0.96 to 1.11, have a mean of 1.04, and a girders in three existing bridges. Results from the proposed
COV of 4%. The predictions from the other methods differ evaluation procedure were found to compare well with
much more from the shear strengths obtained from results from an MCFT fiber model Response-2000.
Response-2000.
ACKNOWLEDGMENTS
CONCLUSIONS This project was supported by the Bridge Engineering Branch of the
This paper presents a shear strength evaluation procedure Ministry of Transportation of British Columbia, Canada.
for structural concrete girders that contains at least minimum
stirrups. The procedure provides information about the shear NOTATION
strength at three failure modes: first yielding of stirrups, A p, A = area of prestressed, nonprestressed flexural tension reinforcement,
concrete crushing after stirrup yielding, and longitudinal respectively
Asw, Apw = area of prestressed, nonprestressed longitudinal reinforcement
reinforcement yielding, as well as includes a check on centered in web, respectively
concrete crushing before stirrup yielding. A complete Atf = area of concrete in tension surrounding flexural tension
summary of the procedure is given in the Appendix. The reinforcement

ACI Structural Journal/July-August 2009 425

Av = area of transverse reinforcement (stirrups) spaced at s 7. MacGregor, J. K., and Wight, J. K., Reinforced Concrete Mechanics
bw = width of shear area (web) and Design, fourth edition, Prentice Hall, Upper Saddle River, NJ, 2004,
d = depth from compression face to centroid of flexural tension 1132 pp.
reinforcement 8. Collins, M. P.; Mitchell, D.; Adebar, P.; and Vecchio, F. J., “A General
dnv = depth of uniform compression stress nv over diagonally Shear Design Method,” ACI Structural Journal, V. 93, No. 1, Jan.-Feb.
cracked web, ≤dv 1996, pp. 36-45.
dpw = depth from compression face to centroid of prestressing 9. Collins, M. P., and Rahal, K. N., “Background to General Method of
tendons in web Shear Design in 1994 CSA-A23.3 Standard,” Canadian Journal of Civil
dv = depth of uniform shear stress, may be taken as 0.9d Engineering, V. 26, No. 6, 1999, pp. 827-839.
Ep,Es = modulus of elasticity of prestressed, nonprestressed reinforce- 10. Bentz, E. C., and Collins, M. P., “Development of the 2004 Canadian
ment, respectively Standards Association (CSA) A23.3 Shear Provisions for Reinforced
fc′ = specified compressive strength of concrete Concrete,” Canadian Journal of Civil Engineering, V. 33, No. 5, 2006,
fcx , fcz = concrete normal stress in x- and z-directions, respectively pp. 521-534.
fp = effective stress in prestressing tendons 11. Bentz, E. C.; Vecchio, F. J.; and Collins, M. P., “Simplified Modified
fpr = stress in prestressing tendons at maximum resistance Compression Field Theory for Calculating Shear Strength of Reinforced
fsx , fsz = stress in x- and z-directions reinforcement, respectively Concrete Elements,” ACI Structural Journal, V. 103, No. 4, July-Aug.
fx , fz = applied normal stress in x- and z-directions, respectively 2006, pp. 614-624.
fy = yield strength of reinforcement
12. Esfandiari, A., “Shear Strength of Structural Concrete Members
jd = internal flexural lever-arm, may be taken as 0.9d except in
Using a Uniform Shear Element Approach,” PhD thesis, Department of
Eq. (24)
Civil Engineering, University of British Columbia, Vancouver, BC,
jdpw = internal flexural lever-arm of prestressing tendons in web
Canada, Apr. 2009, 333 pp.
M = bending moment at section of interest
Nv, Nvc = resultant of nv , nvc stress assumed uniform over effective 13. Hawkins, N. M.; Kuchma, D. A.; Mast, R. F.; Marsh, M. L.; and
depth dnv Reineck, K. H., “Simplified Shear Design of Structural Concrete
nv = axial compression stress in concrete required to resist Members,” NCHRP Report 549, American Association of State Highway
applied shear and Transportation Officials, Washington, DC, 2005.
nv* = axial compression stress required to resist applied shear at 14. Kim, K. S., “Shear Behavior of Reinforced Concrete Beams and
location of diagonal cracks Prestressed Concrete Beams,” PhD thesis, University of Illinois, Urbana,
nvc = maximum nv as limited by reserve capacity of longitudinal IL, July 2004, 474 pp.
reinforcement 15. Shahawy, M. A., and Batchelor, B., “Shear Behavior of Full-Scale
nvo = predicted value of nv at εx = 0 Prestressed Concrete Girders: Comparison with AASHTO Specifications,”
s = spacing of transverse reinforcement (stirrups) PCI Journal, V. 41, No. 3, 1996, pp. 48-62.
V = total shear resistance 16. Rahal, K. N., and Al-Shaleh, K. S., “Minimum Transverse Reinforcement
Vc, Vs = shear resistance attributed to concrete, stirrups, respectively in 65 MPa Concrete Beams,” ACI Structural Journal, V. 101, No. 6, Nov.-
Vp = shear force resisted by inclined prestressing tendons Dec. 2004, pp. 872-878.
v = total shear stress 17. Mphonde, A. G., and Frantz, G. C., “Shear Tests of High- and Low-
vc, vs = shear stress attributed to concrete, stirrups, respectively Strength Concrete Beams with Stirrups,” High-Strength Concrete, SP-87,
α = concrete tension strength factor American Concrete Institute, Farmington Hills, MI, 1985, pp. 179-196.
β = concrete shear contribution factor 18. Ma, Z.; Tadros, M. K.; and Baishya, M., “Shear Behavior of
Δnv = predicted rate of change of nv per unit εx Pretensioned High-Strength Concrete Bridge I-Girders,” ACI Structural
Δθ = predicted rate of change of θ per unit εx Journal, V. 97, No. 1, Jan.-Feb. 2000, pp. 185-192.
εx = average longitudinal strain over depth of member 19. ACI Committee 318, “Building Code Requirements for Structural
εy = yield strain of transverse reinforcement (stirrups) Concrete (ACI 318-05) and Commentary (318R-05),” American Concrete
εz = strain of transverse (z-direction) reinforcement Institute, Farmington Hills, MI, 430 pp.
γ = shear strain
λ = ratio jdpw /jd
θ = angle of inclination of principle compression stress (diagonal APPENDIX: SUMMARY OF PROPOSED
cracks) PROCEDURE
θo = predicted value of θ at εx = 0
ρx = ratio of distributed longitudinal (x-direction) reinforcement
area to concrete area Concrete crushing after
ρz = ratio of transverse (z-direction) reinforcement area to Yielding of stirrups stirrup yielding
concrete area (1) β = 2.17(1.6 – 300εy) > 2.17 psi
(1) β from Eq. (10)
β = 0.18(1.6 – 300εy) > 0.18 MPa
REFERENCES (2) θo and Δθ from Eq. (6) and (7) (2) θo and Δθ from Eq. (8) and (9)
1. Vecchio, F. J., and Collins, M. P., “The Modified Compression Field
Theory for Reinforced Concrete Elements Subjected to Shear,” ACI (3) If θo > 23 degrees, Δnv and nvo
JOURNAL, Proceedings V. 83, No. 2, Mar.-Apr. 1986, pp. 219-231. (3) Δnv and nvo from Eq. (13) from Eq. (13) and (14)
2. “AASHTO LRFD Bridge Design Specifications,” fourth edition, and (14) If θo ≤ 23 degrees, Δnv and nvo
American Association of State and Highway Transportation Officials, from Eq. (15) and (16)
Washington DC, 2007, 1518 pp. (4) εx from Eq. (21) (4) εx from Eq. (21)
3. “AASHTO LRFD Bridge Design Specifications,” 2008 interim revisions,
(5) If Eq. (22) satisfied, multiply εx (5) If Eq. (22) satisfied, multiply εx
American Association of State and Highway Transportation Officials,
by 2 by 2
Washington DC, 2008, 652 pp.
4. “Canadian Highway Bridge Design Code (CAN/CSA-S6-06),” Canadian (6) θ from Eq. (5) (6) θ from Eq. (5)
Standards Association, Toronto, ON, Canada, 2006, 733 pp. (7) Vyield from Eq. (1) (7) Vcrush from Eq. (1)
5. Bentz, E. C., “Sectional Analysis of Reinforced Concrete Members,”
PhD thesis, Department of Civil Engineering, University of Toronto, Governing stirrup yielding mode: (8) Vmax = max(Vyield ,Vcrush)
Toronto, ON, Canada, 2000. Check yielding of longitudinal reinforcement: (9) Vn ≤ Vlong, where
6. Esfandiari, A., and Adebar, P., “Shear Strength Evaluation of Structural Vlong from Eq. (23) to (25).
Concrete Girders with Less Than Minimum Stirrups,” 2008 Concrete
Bridge Conference, St. Louis, MO, May 2008, 20 pp. Check concrete crushing before stirrup yielding: (10)Vn ≤ 0.25fc′ bw dv + Vp

426 ACI Structural Journal/July-August 2009