Strength Prediction for Discontinuity Regions by Softened

Strut-and-Tie Model
Shyh-Jiann Hwang1 and Hung-Jen Lee2
Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

Abstract: Discontinuities caused by abrupt changes in cross-sectional dimensions or by concentrated loads result in discontinuity
regions due to disturbance in the flow of internal forces. A simplified method, based on the softened strut-and-tie model, for determining
the shear strength of discontinuity regions failing in diagonal compressions is proposed in this paper. Strength predictions of the resulting
expressions correlate well with the 449 test results of deep beams, corbels, squat walls, and beam-column joints available from the
literature. The proposed method incorporates the shear resisting mechanisms as postulated by the softened strut-and-tie model, and it is a
function of the concrete strength, horizontal shear reinforcement, vertical shear reinforcement, and geometrical configuration of the
discontinuity regions. A numerical example is included for illustration.
DOI: 10.1061/共ASCE兲0733-9445共2002兲128:12共1519兲
CE Database keywords: Compression; Discontinuities; Concrete, reinforced; Shear; Strength.

Introduction given situation, many strut-and-tie models are possible and there
is no unique solution 共ASCE-ACI Committee 445 1998兲.
There are many situations in the strength prediction of reinforced A softened strut-and-tie model, satisfying equilibrium, com-
concrete members where discontinuities such as abrupt changes patibility, and constitutive laws of cracked reinforced concrete,
in cross-sectional dimensions or the presence of concentrated has been proposed for determining the shear strengths of beam-
loads or reactions cause disturbances in the flow of internal column joints 共Hwang and Lee 1999, 2000兲, deep beams 共Hwang
forces. In the discontinuity regions, the shear stresses are no et al. 2000a兲, corbels 共Hwang et al. 2000b兲, and squat walls
longer uniform over the depth of the member. The main force 共Hwang et al. 2001兲. In order to satisfy the three principles of
flow is through a major compression diagonal between load and mechanics of materials rigorously, a successful prediction of the
support, and the failure in this region is usually governed by previously proposed model 共Hwang and Lee 1999, 2000兲 lies on
crushing the diagonal compressive strut as shown in Fig. 1. elaborate calculations, and hence the distinct advantage of sim-
The strut-and-tie model is particularly useful in the strength plicity of the strut-and-tie model is lost.
prediction of discontinuity regions because the flow of forces can On the basis of the softened strut-and-tie model 共Hwang and
be easily visualized by identifying the discontinuity regions with Lee 1999, 2000兲 derived previously, the present study concen-
compressive struts representing the flow of concentrated compres- trates on developing a simple predicting procedure, which deals
sive stresses in the concrete and tension ties representing the re- with the strength prediction of reinforced concrete discontinuity
inforcing steel. This strut-and-tie model 共or truss model兲, pio- regions failing in diagonal compressions without getting lost in
neered by Ritter 共1899兲 and Mörsch 共1909兲, has been greatly trivialities.
advanced during the past decades 共ASCE-ACI Committee 445
1998兲.
Although the strut-and-tie modeling is rational, it is often Research Significance
termed as an approximate approach with undetermined accuracy
共Hsu 1996兲. In the strut-and-tie method, the stresses are usually The current design provisions on deep beams, corbels, squat
determined by equilibrium conditions alone, while the strain com- walls, and beam-column joints of the ACI 318 building code 共ACI
patibility conditions are neglected. Without the compatibility con- 1999兲 are derived from empirical test data, and consequently
straints, engineers are free to choose the efficient strut-and-tie these provisions result in different formats. Since they all repre-
models that satisfy the equilibrium conditions. Therefore, for any sent the discontinuity regions failing in diagonal compressions
共Fig. 1兲, their design provisions should be unified on a rational
1
Professor, Dept. of Construction Engineering, National Taiwan Univ. basis. According to the softened strut-and-tie model 共Hwang and
of Science and Technology, Taipei, Taiwan 10672, ROC. Lee 1999, 2000兲, this paper presents a simplified approach that is
2
Assistant Professor, Dept. of Construction Engineering, National a useful and practical tool for determining the shear strength of
Yunlin Univ. of Science and Technology, Yunlin, Taiwan 64045, ROC. the discontinuity regions failing in diagonal compressions.
Note. Associate Editor: Joseph M. Bracci. Discussion open until May
1, 2003. Separate discussions must be submitted for individual papers. To
extend the closing date by one month, a written request must be filed with
the ASCE Managing Editor. The manuscript for this paper was submitted
Softened Strut-and-Tie Model
for review and possible publication on March 23, 2001; approved on
March 15, 2002. This paper is part of the Journal of Structural Engi- Fig. 2 shows the strut-and-tie modeling of an interior beam-
neering, Vol. 128, No. 12, December 1, 2002. ©ASCE, ISSN 0733-9445/ column joint under earthquake-induced forces. Bond deterioration
2002/12-1519–1526/$8.00⫹$.50 per page. along the beam and column reinforcement is assumed, and the

JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002 / 1519

J. Struct. Eng., 2002, 128(12): 1519-1526

Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Reinforced concrete discontinuity regions failing in diagonal

compressions

principal stress is concentrated along the diagonal strut to cause

shear failure. The concentrated flow of stresses is modeled as the
planar truss shown in Fig. 2, which is composed of diagonal,
horizontal, and vertical mechanisms 共Hwang and Lee 1999,
Fig. 3. Forces in concrete struts
2000兲. Considering the horizontal shear force V h as an external
load 共Fig. 2兲, the other three resultants from the compression
zones can be treated as the reactions to restrain the planar truss in Fh Fv
C d ⫽⫺D⫹ ⫹ (4)
stable equilibrium. The equilibrium of the entire truss in Fig. 2 cos ␪ sin ␪
leads to the following expressions:
where D⫽compression force in the diagonal strut, negative in
V v ⫽C d sin ␪ (1) compression; and F h and F v ⫽tension forces in the horizontal and
vertical ties, respectively.
V h ⫽C d cos ␪ (2) Eq. 共4兲 expresses the components of the diagonal compression
Vv ᐉv to be apportioned to the diagonal, horizontal, and vertical mecha-
⫽ ⫽tan ␪ (3) nisms. It is further assumed that the ratios of the diagonal com-
Vh ᐉh
pression C d assigned among the three mechanisms are defined as
where V v ⫽vertical shear forces; C d ⫽diagonal compression;
Fh Fv
␪⫽angle of inclination of the diagonal compression with respect ⫺D: : ⫽R d :R h :R v (5)
to the horizontal h axis; and ᐉ h and ᐉ v ⫽internal lever arms of the cos ␪ sin ␪
vertical and horizontal shear couples 共Hwang and Lee 1999, where R d , R h , and R v ⫽ratios of the diagonal compression re-
2000兲, respectively. sisted by the diagonal, horizontal, and vertical mechanisms, re-
Fig. 3 presents the forces in concrete struts, and the diagonal spectively.
compression to be resisted by the strut-and-tie model is found to The truss in Fig. 2 is statically indeterminate to the second
be degree, therefore, the values of R d , R h , and R v cannot be deter-
mined by equilibrium alone. The diagonal compression C d will be
distributed among the resisting mechanisms in proportion to their
relative stiffnesses. Since Schäfer 共1996兲, and Jennewein and
Schäfer 共1992兲 provided two additional equations on the stiffness
ratios among the resisting mechanisms, the values of R d , R h , and
R v can be fully defined.
According to Schäfer 共1996兲 and Jennewein and Schäfer
共1992兲, the stiffness ratio between the elastic horizontal tie and
the diagonal strut for transferring the diagonal compression is
␥ h /(1⫺␥ h ), where ␥ h ⫽fraction of diagonal compression trans-
ferred by the horizontal tie in the absence of the vertical tie.
Based on the study of the principal stress pattern from a linear-
elastic finite element analysis 共Schäfer 1996兲, the value of ␥ h is
defined as 关adopted in the CEB-FIP Model Code 1990 共1993兲,
Sect. 6.8.2.2.兴
␥ h ⫽ 共 2 tan ␪⫺1 兲 /3 for 0⭐␥ h ⭐1 (6)
The relative stiffness ratio between the horizontal and the diago-
nal mechanisms should be the same with or without the partici-
Fig. 2. Strut-and-tie modeling of interior joint
pation of the vertical mechanism, and that is

1520 / JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002

J. Struct. Eng., 2002, 128(12): 1519-1526

 R h :R d ⫽␥ h : 共 1⫺␥ h 兲 (7) ures of the discontinuity regions due to diagonal compressions are
also governed by the above phenomenon of softening of concrete
The stiffness ratio between the elastic vertical tie and the di- 共Hwang and Lee 1999, 2000兲.
agonal strut for transferring the diagonal compression is ␥ v /(1 In the present study, the concrete softened laws are taken as
⫺␥ v ), where ␥ v ⫽fraction of diagonal compression carried by the given by Zhang and Hsu 共1998兲. The ascending branch of the
vertical tie in the absence of the horizontal tie. The value of ␥ v is softened stress-strain curve of the cracked concrete in compres-
defined as sion is given as
␥ v ⫽ 共 2 cot ␪⫺1 兲 /3 for 0⭐␥ v ⭐1
Similarly, the stiffness ratio between the vertical and the diagonal
(8)
␴ d ⫽⫺␨ f ⬘c 2 冋冉 冊 冉 冊册
⫺⑀ d
␨⑀ 0
⫺
⫺⑀ d
␨⑀ 0
2
for
⫺⑀ d
␨⑀ 0
⭐1 (17)
mechanisms can be defined as
5.8 1 0.9
Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

R v :R d ⫽␥ v : 共 1⫺␥ v 兲 ␨⫽ ⭐ (18)
(9) 冑 f c⬘ 冑1⫹400⑀ r 冑1⫹400⑀ r
It is useful to set the sum of R d , R h , and R v equal to unity. where ␴ d ⫽average principal stress of concrete in the d direc-
R d ⫹R h ⫹R v ⫽1 (10) tion, negative in compression; ␨⫽softening coefficient;
f ⬘c ⫽compressive strength of a standard concrete cylinder in units
By solving Eqs. 共7兲, 共9兲, and 共10兲, the values of R d , R h , and of MPa; ⑀ d and ⑀ r ⫽average principal strains in the d and r direc-
R v can be obtained as tions, respectively; and ⑀ 0 ⫽concrete cylinder strain correspond-
共 1⫺␥ h 兲共 1⫺␥ v 兲 ing to the cylinder strength f ⬘c .
R d⫽ (11) If the maximum compressive stress exerting on the nodal zone
1⫺␥ h ␥ v
⫺␴ d,max is less than the capacity of the cracked concrete ␨ f ⬘c , the
␥ h 共 1⫺␥ v 兲 diagonal compression could be increased. The shear strength of
R h⫽ (12) the discontinuity region is obtained when the acting stress
1⫺␥ h ␥ v
⫺␴ d,max exceeds the concrete capacity ␨ f ⬘c .
␥ v 共 1⫺␥ h 兲 In order to calculate the value of ␨, the principal tensile strain
R v⫽ (13) ⑀ r should be determined. It is assumed that the two-dimensional
1⫺␥ h ␥ v
membrane elements of the discontinuity regions should satisfy
All the member forces of the truss shown in Fig. 2 can be deter- Mohr’s circular compatibility. The equation of the first strain in-
mined for any specified diagonal compression C d with the aid of variant is used
␥ h , ␥ v , R d , R h , and R v .
The failure of the compression strut is defined as the crushing ⑀ r ⫹⑀ d ⫽⑀ h ⫹⑀ v (19)
of concrete at the end of the diagonal strut immediately adjoining where ⑀ h and ⑀ v ⫽averaged normal strains in the h and v direc-
the nodes 共Fig. 2兲. In order to check whether the failure stress is tions, respectively.
reached, the area of the nodal zone which bears the diagonal If the stress-strain relationship of the shear reinforcement is
pressure must be estimated. The effective area of the diagonal assumed to be elastic-perfectly plastic, the values of ⑀ h and ⑀ v can
strut A str is defined as be estimated from the forces of the tension ties.
A str⫽a s ⫻b s (14) F h ⫽A th E s ⑀ h ⭐F yh (20)
where a s ⫽depth of the diagonal strut; and b s ⫽width of the diag- F v ⫽A t v E s ⑀ v ⭐F y v (21)
onal strut 共Hwang and Lee 1999, 2000兲.
where A th and A t v ⫽areas of the horizontal and vertical ties
The maximum compressive stress ␴ d,max resulting from the
共Hwang and Lee 1999, 2000兲, respectively; E s ⫽elastic modulus
summation of the compressive forces from the diagonal, flat, and
of the steel bars; and F yh and F y v ⫽yielding forces of the hori-
steep struts 共Fig. 3兲 on the nodal zone can be estimated as
zontal and vertical ties, respectively.

⫺␴ d,max⫽
1
A str冋⫺D⫹
Fh
cos ␪ f
cos共 ␪⫺␪ f 兲
The strength analysis of the extent of softening involves five
unknowns, viz. ␨, ⑀ d , ⑀ r , ⑀ h , and ⑀ v . There are five equations
given by the strain compatibility and the constitutive laws of con-

⫹
Fv
sin ␪ s
cos共 ␪ s ⫺␪ 兲 册 (15)
crete and steel, i.e., Eqs. 共17兲 to 共21兲. The solution can be ob-
tained through iteration procedures, and more details can be
found in Hwang and Lee 共1999, 2000兲.
where ␴ d,max is negative in compression; and ␪ f and ␪ s ⫽angles of
inclination of h axis with respect to flat and steep struts, respec-
tively. Incorporating the relations 2 tan ␪ f ⫽tan ␪ and tan ␪ s Derivation of Simplified Equations
⫽2 tan ␪ 共Figs. 2 and 3兲, Eq. 共15兲 becomes Design equations should enable the engineers to grasp the essen-

⫺␴ d,max⫽
1
A str冋⫺D⫹
Fh
cos ␪
1⫺
sin ␪
2 冉 2
冊 tials of phenomenon without carrying through tedious analyses.
The capacity of the designer in both understanding and solving
problems can be enhanced by reducing the number of variables

⫹
Fv
sin ␪ 冉
1⫺
cos2 ␪
2 冊册 (16)
involved in the design equations. The above-mentioned strut-and-
tie model should therefore be simplified so as to keep its number
of variables as small as possible and still be capable of adequately
Cracked reinforced concrete in compression has been observed describing the problems.
to exhibit lower strength than uniaxially compressed concrete A case study on the shear strength prediction of corbels
共Vecchio and Collins 1993; Zhang and Hsu 1998兲. The shear fail- 共Hwang et al. 2000b兲 had shown that a bilinear relationship exists

JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002 / 1521

J. Struct. Eng., 2002, 128(12): 1519-1526

 between the shear strength and the horizontal reinforcement. The
concrete strength appears to set an upper limit value of horizontal
tie, below which 共underreinforced兲 the horizontal hoops reach
yielding before failure, at which 共balanced point兲 the horizontal
hoops reach yielding at failure, and above which 共overreinforced兲
the horizontal hoops do not yield 共Hwang et al. 2000b兲. For the
under-reinforced corbels, the shear capacities are increased with
increasing horizontal reinforcement at a greater rate 共Hwang et al.
2000b兲 because the horizontal hoops are effective in augmenting
the force of tension tie F h . For the overreinforced corbels, the
shear capacities are increased with increasing horizontal rein-
forcement at a much lower rate 共Hwang et al. 2000b兲 because the
Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

additional horizontal reinforcement beyond the balanced amount

is only effective in retarding the softening effect of concrete. Fig. 4. Diagonal mechanism
The above observations revealed that the beneficial effect of
the additional tension tie beyond the balanced amount on shear
strength could be neglected. The discontinuity region with the yielding strains in the softened strut-and-tie model 共Hwang and
balanced amount of tension ties is almost achieving its maximum Lee 1999, 2000兲. From past experience 共Hwang and Lee 1999,
capacity of diagonal compression. Moreover, the linear relation- 2000兲, the variable ⑀ d can be relegated to a constant of ⫺0.001
ship between the shear strength and the amount of tension tie for strength estimation. Therefore the value of ⑀ r is taken as 0.005
indicates that the technique of linear interpolation can be used for 共0.002⫹0.002⫹0.001兲 as per Eq. 共19兲. As a result, the softening
the underreinforced discontinuity regions for strength prediction. coefficient ␨ can be approximated by Eq. 共18兲 as
The role of the ties in increasing shear strength can be ex-
plained as follows. If a discontinuity region contains the tie rein- ␨⬇3.35/冑 f ⬘c ⭐0.52 (24)
forcement, the diagonal compression will be carried by additional
load paths instead of the diagonal strut alone. This will activate
Strut-and-Tie Index
more concrete for shear resistance, and thus enhance the shear
strength. A factor representing the beneficial effect of the tie force There are four possible combinations of resisting mechanisms for
on the shear strength can take the form discontinuity regions subjected to diagonal compressions. They
are the diagonal mechanism, the diagonal plus horizontal mecha-
Cd nisms, the diagonal plus vertical mechanisms, and the complete
K⫽
⫺␴ d,max⫻A str mechanisms. The strut-and-tie index for each combination is de-
fined separately as follows.
Fh Fv If a discontinuity region is not detailed with any tie reinforce-
⫺D⫹ ⫹
cos ␪ sin ␪ ment, the diagonal compression C d is carried only by the diagonal

冉 冊 冉 冊
⫽ ⭓1 (22)
Fh sin2 ␪ Fv cos2 ␪ strut, as shown in Fig. 4. Recalling Eq. 共22兲, we can state the
⫺D⫹ 1⫺ ⫹ 1⫺ diagonal strut index K d as
cos ␪ 2 sin ␪ 2
where K is defined as the strut-and-tie index. K d ⫽ 共 ⫺D 兲 / 共 ⫺D 兲 ⫽1 (25)
The numerator of Eq. 共22兲 is the diagonal compression as A strut-and-tie model consisting of diagonal and horizontal
stated by Eq. 共4兲, and the denominator is the crushing force as mechanisms is depicted in Fig. 5 to realize the flow of forces in
defined by Eq. 共16兲. As shown in the denominator of Eq. 共22兲, the the disturbed region. As shown in Fig. 5共a兲, in the presence of
existence of tension tie forces reduces the compressive stress, horizontal reinforcement, more flat substruts are developed, and
⫺␴ d,max , thus enhancing the shear strength. internal stress flows are deviated from the direction of C d
In the present study, the nominal diagonal compressive 共Rogowsky and MacGregor 1986兲.
strength C d,n is estimated as For the case of the discontinuity region in Fig. 5 with suffi-
C d,n ⫽K␨ f c⬘ A str (23) cient horizontal reinforcement, the concrete strut would reach its

Eq. 共23兲 had been suggested by Zhang and Jirsa 共1982兲 but with
empirically determined K and ␨. The frame of design equation
outlined by Zhang and Jirsa 共1982兲 will be followed, whereas the
expressions of K and ␨ will be derived theoretically conforming
to the shear resisting mechanisms as postulated by the softened
strut-and-tie model.

Approximation of Softening Effect

The extent of softening of the concrete is directly related to the

value of the principal tensile strain ⑀ r 共Vecchio and Collins 1993;
Zhang and Hsu 1998兲. It was recommended that the value of ⑀ r be
limited at the strain level at which the yielding of the reinforce-
ment crossing the crack occurs 共Vecchio and Collins 1993兲. With
Fig. 5. Diagonal and horizontal mechanisms
this aim in mind, the values of ⑀ h and ⑀ v are limited by their

1522 / JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002

J. Struct. Eng., 2002, 128(12): 1519-1526

Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 6. Complete mechanisms

strength while the horizontal tie remains in the elastic range. The
additional contribution of the sufficient horizontal tie for the di-
agonal compressive strength can be estimated by Eq. 共22兲 as 关Fig.
Fig. 7. Strut-and-tie indexes for overreinforced cases
5共b兲兴

共 1⫺␥ h 兲 ⫹␥ h

冉 冊
K̄ h ⫽ ⭓1 (26)
sin2 ␪ K̄⫽K d ⫹ 共 K̄ h ⫺1 兲 ⫹ 共 K̄ v ⫺1 兲 ⫽K̄ h ⫹K̄ v ⫺1 (31)
共 1⫺␥ h 兲 ⫹␥ h 1⫺
2 Similarly, the value of K can be estimated approximately as
where K̄ h ⫽horizontal tie index with sufficient horizontal rein- K⫽K d ⫹ 共 K h ⫺1 兲 ⫹ 共 K v ⫺1 兲 ⫽K h ⫹K v ⫺1 (32)
forcement. Eq. 共26兲 can be further simplified as
The correct figures and the approximated values of the strut-
1 and-tie indexes for overreinforced cases are plotted in Fig. 7. The
K̄ h ⬇ (27)
1⫺0.2共 ␥ h ⫹␥ 2h 兲 values of K̄ h and K̄ v are approximated reasonably well, and the
approximations of K̄ are overestimated with a noticeable differ-
The balanced amount of the horizontal tie force F̄ h , at which ence at ␪⫽45° 共Fig. 7兲.
the horizontal tie reaches yielding at failure for the case shown in
Fig. 5共b兲, can be defined as
F̄ h ⫽␥ h ⫻ 共 K̄ h ␨ f ⬘c A str兲 ⫻cos ␪ (28) Comparison with Experiments

For the under-reinforced cases, the horizontal tie index K h can The estimations of the proposed simple method are compared
be approximated by linear interpolation using the balanced with the measured peak strength for 449 specimens tested in the
amount of the horizontal tie force. United States, Canada, England, Japan, New Zealand, Australia,
and Taiwan. The test data include 123 deep beams, 178 corbels,
K h ⫽1⫹ 共 K̄ h ⫺1 兲 ⫻F yh /F̄ h ⭐K̄ h (29) 62 squat walls, 49 exterior beam-column joints, and 37 interior
The presence of vertical reinforcement results in steeper sub- beam-column joints. The same test data were used to examine the
struts, so that more concrete is summoned for shear resistance. applicability of the softened strut-and-tie model to deep beams
The set of equations for vertical tie is the same as Eqs. 共26兲–共29兲, 共Hwang et al. 2000a兲, corbels 共Hwang et al. 2000b兲, squat walls
except that all the subscripts h are replaced by v and cos ␪ and 共Hwang et al. 2001兲, exterior joints 共Hwang and Lee 1999兲, and
sin ␪ are interchanged. interior joints 共Hwang and Lee 2000兲. Only the test specimens of
The flow of forces and the resulting strut-and-tie idealization beam-column joints reported as failing in flexure 共Hwang and Lee
for a discontinuity region detailed with horizontal and vertical tie 1999, 2000兲 are excluded in this comparison. The brief descrip-
reinforcement are exhibited in Fig. 6. Owing to the presence of tions of these test specimens and the related references can be
the horizontal and vertical ties, the flow of forces is turned away found in Hwang and Lee 共1999, 2000兲 and Hwang et al. 共2000a,b,
from the diagonal direction 关Fig. 6共a兲兴, which will certainly re- 2001兲.
duce the crushing pressure. The maximum beneficial effect of the Accuracy for the proposed procedures is gauged in terms of a
sufficient tension ties in increasing the shear strength can be cal- strength ratio, which is defined as the ratio of the measured to the
culated by Eq. 共22兲 as 关Fig. 6共b兲兴 computed strengths. Fig. 8 shows satisfactory results for the com-
parison between the measured and computed strengths. The aver-
R d ⫹R h ⫹R v age strength ratios for all the members are close to 1.20, and the

冉 冊 冉 冊
K̄⫽ ⭓1 (30)
sin2 ␪ cos2 ␪ coefficients of variation are all close to 0.20.
R d ⫹R h 1⫺ ⫹R v 1⫺ In Fig. 8, the depth of the diagonal strut a s of the exterior joint
2 2
includes only the contribution from the compression zone of the
where K̄⫽strut-and-tie index with sufficient horizontal and verti- column a c and neglects the effect of the compression zone of the
cal ties. beam a b 共Hwang and Lee 1999兲. Due to the better conditions of
In order to simplify the whole process, it is recommended that joint confinement provided by the framing beams, the a s of the
the value of K̄ be approximated by simple addition. interior joint in Fig. 8 includes the beneficial effect of a beam by

JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002 / 1523

J. Struct. Eng., 2002, 128(12): 1519-1526

Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 8. Experimental verifications

taking a b ⫽h b /5 共where h b is the beam depth兲. Table 1 shows the estimations from one situation to another. Therefore, it can be
further comparison of strength ratios obtained by a s ⫽a c with easily evaluated by or incorporated with the experience of prac-
those estimated by a s ⫽ 冑a 2c ⫹(h b /5) 2 . For completeness, the tice to facilitate decisions regarding strength design.
joints failing in flexure included in the earlier studies 共Hwang and A numerical example that illustrates the application of the pro-
Lee 1999, 2000兲 are also computed in Table 1. posed equations is included in Appendix I.
The proposed simple method is compared with the theoretical
solutions 共general method兲 in Hwang and Lee 共1999, 2000兲 and
Hwang et al. 共2000a,b, 2001兲. The shear strengths of the total 449 Conclusions
specimens listed in Hwang and Lee 共1999, 2000兲 and Hwang
et al. 共2000a,b, 2001兲 are computed by both the simple and the Integrated and simplified equations, which incorporates the shear
general methods as presented in Fig. 9. Good accuracy is obtained resisting mechanisms as postulated by the softened strut-and-tie
for this comparison as indicated in Fig. 9. The predictions of the model, have been formulated to predict the shear strength of dis-
proposed simple method are very close but slightly conservative continuity regions of reinforced concrete members failing in di-
to the theoretical solutions 共Fig. 9兲. agonal compressions. According to the available test results in the
The conservatism of the simplified method is mainly attribut- literature and their comparison with the proposed model, the fol-
able to the approximation of the softening coefficient ␨. The av- lowing conclusions can be made.
erage value of ␨ computed by the general method in Fig. 9 is 1. Eq. 共23兲 is recommended for predicting the diagonal com-
0.537, and the coefficient of variation is 13%. The average value pressive strength of the discontinuity regions. The formula
of ␨ estimated by the simplified method is 0.493 with the coeffi- expresses explicitly the contributions of the tie reinforce-
cient of variation of 10%. Both numbers of 0.537 and 0.493 are ment, concrete strength with softening effect, and cross-
close to the value of 0.51 共0.85⫻0.60), which is adopted by the sectional area of strut.
ACI 318 building code 共ACI 2002兲 as the strength coefficient of 2. The extent of softening of the concrete is expressed by the
the concrete strut with insufficient reinforcement. softening coefficient ␨, and the beneficial effect of the ten-
The proposed method incorporates the shear resisting mecha- sion ties on the shear strength is represented by the strut-and-
nisms as postulated by the softened strut-and-tie model, and it tie index K. In order to facilitate the strength estimation of
provides a rational basis to unify the strength estimations on deep discontinuity regions, the expressions of ␨ and K are greatly
beams, corbels, squat walls, and beam-column joints. The essen- simplified as defined in Eqs. 共24兲 and 共32兲, respectively.
tial elements of the model suggested are apparently that of sim- 3. The method proposed herein was compared with the mea-
plicity and uniqueness. The simplified method is a convenient sured peak strength for 449 specimens of deep beams, cor-
way of converting the strength prediction of discontinuity regions bels, squat walls, and beam-column joints available in the
into a manageable form. The model maintains consistency in its literature, and satisfactory correlation was found.

Table 1. Statistics of Strength Ratios of Beam-Column Joints

C d,test /C d,calc
a s ⫽a c a a s ⫽ 冑a c2 ⫹(h b /5) 2 a
Joint type Failure mode No. of Specimen Avg/Cov Avg/Cov
Exterior Beam flexure 14 0.98/0.24 0.81/0.24
Joint shear 49 1.20/0.21 0.97/0.17
Interior Beam flexure 19 1.09/0.35 0.91/0.35
Joint shear 37 1.45/0.21 1.28/0.17
a
a s ⫽a c or a s ⫽ 冑a c2 ⫹(h b /5) 2 is assumed for failure modes of F1, J1, and J2 共Hwang and Lee 1999, 2000兲.

1524 / JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002

J. Struct. Eng., 2002, 128(12): 1519-1526

 A str⫽a s ⫻b s ⫽168⫻500⫽84,000 mm2
F yh ⫽2⫻4⫻71⫻414⫻10⫺3 ⫽235 kN
F y v ⫽2⫻2⫻510⫻414⫻10⫺3 ⫽845 kN

␨⫽3.35/冑 f ⬘c ⫽3.35/冑38⬎0.52; take ␨⫽0.52

Force distribution
␥ h ⫽ 共 2 tan ␪⫺1 兲 /3⫽ 共 2 tan 49°⫺1 兲 /3⫽0.434
␥ v ⫽ 共 2 cot ␪⫺1 兲 /3⫽ 共 2 cot 49°⫺1 兲 /3⫽0.246
Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

Balanced amounts of tie forces

K̄ h ⫽1/关 1⫺0.2共 ␥ h ⫹␥ 2h 兲兴 ⫽1/关 1⫺0.2共 0.434⫹0.4342 兲兴

⫽1.142
Fig. 9. Comparison of simple with general methods for predicting
diagonal compressive strengths K̄ v ⫽1/关 1⫺0.2共 ␥ v ⫹␥ 2v 兲兴 ⫽1/关 1⫺0.2共 0.246⫹0.2462 兲兴

⫽1.065
Acknowledgment
F̄ h ⫽␥ h K̄ h ␨ f ⬘c A str cos ␪⫽0.434⫻1.142⫻0.52⫻38
The writers are grateful to the National Science Council of the
⫻84,000⫻cos 49°⫻10⫺3
Republic of China for financial support under the Project NSC
89-2211-E-011-011. ⫽540 kN

F̄ v ⫽␥ v K̄ v ␨ f ⬘c A str sin ␪⫽0.246⫻1.065⫻0.52⫻38

Appendix I. Illustrative Example
⫻84,000⫻sin 49°⫻10⫺3
Examine the seismic shear strength of the exterior joint shown
schematically in Fig. 10. ⫽328 kN
Demand of joint shear force V jh,u Tie index
V jh,u ⫽1.25⫻414⫻4⫻510⫻10⫺3 K h ⫽1⫹ 共 K̄ h ⫺1 兲 F yh /F̄ h ⫽1⫹ 共 1.142⫺1 兲 ⫻235/540
⫺ 共 1.25⫻414⫻4⫻510⫻368⫻10⫺3 兲 /3500 ⫽1.062⭐K̄ h
⫽945 kN K v ⫽1⫹ 共 K̄ v ⫺1 兲 F y v /F̄ v
Properties of softened strut-and-tie model 共Hwang and Lee
1999兲 ⫽1⫹ 共 1.065⫺1 兲 ⫻845/328⬎K̄ v ; take K v ⫽1.065
␪⫽tan⫺1 共 ᐉ v /ᐉ h 兲 ⫽tan⫺1 共 368/320兲 ⫽49° Result

a s ⫽a c ⫽ 共 0.25⫹0.85N/A g f ⬘c 兲 h c C d,n ⫽ 共 K h ⫹K v ⫺1 兲 ␨ f ⬘c A str⫽ 共 1.062⫹1.065⫺1 兲

⫽ 共 0.25⫹0.85⫻0.10兲 ⫻500⫽168 mm ⫻0.52⫻38⫻84,000⫻10⫺3

⫽1871 kN
␾C d,n cos ␪⫽0.85⫻1,871⫻cos 49°⫽1,043 kN⬎V jh,u

⫽945 kN,OK

References

American Concrete Institute 共ACI兲. 共1999兲. ‘‘Building code requirements

for structural concrete.’’ ACI 318-99 and Commentary (ACI 318R-99),
Farmington Hills, Mich.
American Concrete Institute 共ACI兲. 共2002兲. ‘‘Building code requirements
for structural concrete.’’ ACI 318-02 and Commentary (ACI 318R-02),
Farmington Hills, Mich.
ASCE-ACI Committee 445. 共1998兲. ‘‘Recent approaches to shear design
of structural concrete.’’ J. Struct. Eng., 124共12兲, 1375–1417.
Comité Euro-International du Beton 共CEB兲-Fédération International de la
Precontrainte 共FIP兲. 共1993兲. Model Code 1990, 1993, (MC90), Tho-
Fig. 10. Illustrative example of exterior joint
mas Telford, London.

JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002 / 1525

J. Struct. Eng., 2002, 128(12): 1519-1526

 Hsu, T. T. C. 共1996兲. ‘‘Toward a unified nomenclature for reinforced- Jennewein, M., and Schäfer, K. 共1992兲. ‘‘Standardisierte nachweise von
concrete theory.’’ J. Struct. Eng., 122共3兲, 275–283. 关Also, see discus- häufigen D-Bereichen.’’ DAfStb H. 430, Beuth, Berlin 共in German兲.
sion by Mo, Y. L., and Hsu, T. T. C. 共1997兲. J. Struct. Eng., 123共12兲, Mörsch, E. 共1909兲. Concrete-steel construction, E. P. Goodrich, transla-
1691–1693.兴 tor, McGraw-Hill, New York.
Hwang, S. J., Fang, W. H., Lee, H. J., and Yu, H. W. 共2001兲. ‘‘Analytical Ritter, W. 共1899兲. ‘‘Die bauweise hennebique.’’ Schweizerische Bauzei-
model for predicting shear strength of squat walls.’’ J. Struct. Eng., tung, 33共7兲, 59– 61.
127共1兲, 43–50. Rogowsky, D. M., and MacGregor, J. G. 共1986兲. ‘‘Design of reinforced
concrete deep beams.’’ Concr. Int.: Des. Constr. 8共8兲, 49–58.
Hwang, S. J., and Lee, H. J. 共1999兲. ‘‘Analytical model for predicting
Schäfer, K. 共1996兲. ‘‘Strut-and-tie models for the design of structural
shear strengths of exterior reinforced concrete beam-column joints for
concrete.’’ Notes of Workshop, Dept. of Civil Engineering, National
seismic resistance.’’ ACI Struct. J., 96共5兲, 846 – 857. Cheng Kung Univ., Tainan, Taiwan.
Hwang, S. J., and Lee, H. J. 共2000兲. ‘‘Analytical model for predicting Vecchio, F. J., and Collins, M. P. 共1993兲. ‘‘Compression response of
shear strengths of interior reinforced concrete beam-column joints for cracked reinforced concrete.’’ J. Struct. Eng., 119共12兲, 3590–3610.
Downloaded from ascelibrary.org by Indian Inst Of Technology - Chennai on 01/10/17. Copyright ASCE. For personal use only; all rights reserved.

seismic resistance.’’ ACI Struct. J., 97共1兲, 35– 44. Zhang, L. X. B., and Hsu, T. T. C. 共1998兲. ‘‘Behavior and analysis of 100
Hwang, S. J., Lu, W. Y., and Lee, H. J. 共2000a兲. ‘‘Shear strength predic- MPa concrete membrane elements.’’ J. Struct. Eng., 124共1兲, 24 –34.
tion for deep beams.’’ ACI Struct. J., 97共3兲, 367–376. Zhang, L., and Jirsa, J. O. 共1982兲. ‘‘A study of shear behavior of rein-
Hwang, S. J., Lu, W. Y., and Lee, H. J. 共2000b兲. ‘‘Shear strength predic- forced concrete beam-column joints.’’ PMFSEL Rep. No. 82-1, Dept.
tion for reinforced concrete corbels.’’ ACI Struct. J., 97共4兲, 543–552. of Civil Engineering, Univ. of Texas, Austin, Tex.

1526 / JOURNAL OF STRUCTURAL ENGINEERING / DECEMBER 2002

J. Struct. Eng., 2002, 128(12): 1519-1526