ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 108-S40

Deformation-Based Strut-and-Tie Model for Interior Joints

of Frames Subject to Load Reversal
by Sung-Gul Hong, Soo-Gon Lee, and Thomas H.-K. Kang

This paper presents a strut-and-tie model for reinforced concrete be designed in the elastic state for design purposes. The
(RC) interior beam-column joints to investigate shear strength effect of plastic hinge deformation in the beam on the shear
degradation depending on the deformation of plastic hinges of strength of the beam-column joint has not been clearly
adjacent beams. Two primary factors are considered in the included; however, as the yield penetration of beam bars
development of a deformation-based strut-and-tie model: 1) shear
occurs into the joint, the shear strength damage mechanism
force and moment transmitted from adjacent beams and columns;
and 2) strain distribution of beam longitudinal bars within the is likely to change from the combined truss and diagonal
joint. The bond stress distribution along the beam bars within the strut action to the clear diagonal strut action. Accordingly, it
joint is used to depict appropriate force transfer patterns within is necessary to investigate the relationship between beam
the interior joint. Based on comparisons between the predicted plastic hinge deformation and joint shear strength under
values of strength and deformation capacity and test results from substantial deformation in an effort to propose basic
the literature, the proposed model successfully predicted the concepts that may inspire future design directions for interior
strengths and deformation capacities of interior beam-column beam-column joints. Figure 1 illustrates a deformation-
subassemblages influenced by joint shear failure. based design strategy that is used in this study.
This paper proposes a strut-and-tie model capable of
Keywords: frames; joints; load; strut-and-tie model.
simulating the shear strength degradation of RC interior
beam-column joints dependent on the behavior of adjacent
INTRODUCTION beam plastic hinges. Additionally, the proposed model
Beam-column joints of reinforced concrete (RC) building estimates the attainable shear strength of joints for a given
frames are susceptible to high shear in the event of story drift ratio of beam-column frames and the ductility
earthquakes. In most design practices, one treats beam- limit defined by joint shear failure.
column joints as weak links and attempts to ensure sufficient
shear strengths of joints by providing appropriate detailing
RESEARCH SIGNIFICANCE
and materials to resist the loads transmitted from beams and
columns, even after the formation of beam plastic hinges. Current ACI 318-081 code provisions provide limits on
The complexity and uncertainty of the load-transfer the shear strength of beam-column joints and detailing rules
mechanism in the joint regions generally leads to an to ensure sufficient joint shear strength. The complexity of
ambiguous design methodology; design code provisions load-transfer mechanisms within a joint and the treatment of
such as ACI 318-081 and NZS 3101-952 provide simplified a joint isolated from the adjacent members have led to a
procedures for the design of beam-column joints.3 simplified design with little consideration of the interaction
Shear strengths of RC flexural members are significantly
affected by the post-yield deformation. In other words, the
shear strengths tend to decrease as the displacement ductility
increases. Wight and Sozen4 first acknowledged the reduction
in the shear resistance of RC flexural members due to the
post-yield deformation. Subsequently, the Applied Technology
Council (ATC-65) proposed a conceptual model of shear
strength degradation depending on the displacement
ductility. Previous experimental research6,7 reported that
bond stress loss along flexural bars with yield penetration
and softening of the concrete struts for shear transfer are the
two main influential factors affecting the degradation of
shear strength.
This type of shear strength degradation due to the post-
yield deformation of plastic hinges applies to both flexural Fig. 1—Deformation-based design strategy for beam-column
members and beam-column joints where flexural reinforcement subassemblages.
passes through or terminates. Hakuto et al.7,8 verified the
degradation in the joint shear strength due to the loss of the
bond along beam bars within the joint. As most plastic ACI Structural Journal, V. 108, No. 4, July-August 2011.
MS No. S-2009-304.R2 received April 20, 2010, and reviewed under Institute
hinges of building frames are designed to occur at the beam publication policies. Copyright © 2011, American Concrete Institute. All rights reserved,
ends according to the capacity design concept, a beam- including the 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
column joint has been regarded as an isolated component to 2012 ACI Structural Journal if the discussion is received by January 1, 2012.

ACI Structural Journal/July-August 2011 423

 The truss action for shear transfer in joints requires the
ACI member Sung-Gul Hong is a Professor of architecture at Seoul National
University, Seoul, Korea. He received his BS and MS from Seoul National University development of bond stresses along the beam and column
in 1981 and 1983, respectively, and his PhD from Lehigh University, Bethlehem, PA, bars within the joints. The portion of shear transfer via truss
in 1994. His research interests include strut-and-tie models for bond transfer, shear action decreases as the yield penetration into the joint
strength of reinforced concrete members, shear friction with creep, anchorage of
multiple bars, and deformation of reinforced concrete columns in shear. progresses (after the yielding of the beam bars), resulting in
a loss of bond resistance. The shear strength developed by
Soo-Gon Lee is a Structural Engineer at Samsung Construction Co., Seoul, Korea. the truss action (and arch action) is affected by the stress and
He received his BS, MS, and PhD from Seoul National University in 1998, 2000, and
2004, respectively. His research interests include strut-and-tie models for member strain states in the beam bars. In this study, due to the
deformations and the strength of nodal zones. intended plastic hinge formation in the beams, the strain
ACI member Thomas H.-K. Kang is an Assistant Professor at Seoul National
states in the beam bars at the beam-column interface and
University. He is Secretary of Joint ACI-ASCE Committee 352, Joints and Connections within the joint are selected as the main parameters to
in Monolithic Concrete Structures, and is a member of ACI Committees 335, Composite construct stress fields and an associated strut-and-tie model
and Hybrid Structures; 369, Seismic Repair and Rehabilitation; Joint ACI-ASCE
Committee 423, Prestressed Concrete; and the ACI Collegiate Concrete Council. He
for interior beam-column joints that allows explanation of
received the ACI Wason Medal for Most Meritorious Paper in 2009. His research the shear strength degradation.
interests include a number of subjects relating to the design, repair, and materials of
structural concrete.
Shear forces in interior joints
The forces acting in an interior beam-column joint under
lateral loading are considered. The equilibrium condition
between the joint and adjacent members. This study provides of forces in the horizontal and vertical directions in the joint
a rational model for determining the shear strength of an region is illustrated in Fig. 2, as also detailed by Paulay and
interior beam-column joint depending on the deformation of Priestley.12 The horizontal shear force in the joint core Vjh is
members adjacent to the joint and for estimating the system expressed as
ductility when it is limited by joint shear failure.
V jh = T b1 + C b2 – V c (1)
SHEAR STRENGTH OF INTERIOR
BEAM-COLUMN JOINTS where Tb1 and Cb2 are the tension and compression forces in
To estimate the shear strength of a beam-column joint, it is the beams due to flexure, respectively (refer to Fig. 2); Vc is
generally accepted that internal shear forces are transferred by the column shear force, which is equal to ((My1 + My2/h) +
both the diagonal strut mechanism and the truss mechanism (Vbhc/h)) at beam yielding during the moment reversal.
(for example, Hakuto et al.8). There has been little consensus, Here, My1 and My2 are the positive and negative yield
however, on the determination of the contribution of each moments of the left and right beams, respectively; Vb is the
mechanism. Some researchers (for example, Hakuto et al.7,8 shear force in the beam; hc is the column depth in the
and NZS 3101-952) consider that the joint shear strength is direction of lateral force; and h is the interstory height.
affected by joint transverse reinforcement, whereas other For a beam-column subassemblage subjected to reversed
researchers (for example, Fujii and Morita9 and ACI 318-081) do inelastic deformations, plastic hinges are intended to form at
not consider a term associated with joint transverse the beam ends at the column faces according to the strong
reinforcement in the joint shear strength equation. Hwang column-weak beam design concept. In this state, the forces
and Lee10,11 evaluated the contribution of the truss action of Cb1, which equals Tb1 (refer to Fig. 2), and Tb2 are set as
using two-way arch-truss models with the softening of the As1 fy1 and As2 fy2, respectively. Here, As1 and As2 are the
concrete struts. Based on the two-way arch-truss models, the cross-sectional areas of the tensile reinforcement of the left
force transfer within the joint is once set and then final failure of (bottom) and right (top) beams when subject to lateral loads
the joint is governed by crushing failure of the strut. from the right- to left-hand side, respectively, and fy1 and fy2
denote the specified yield strength of the reinforcing bottom
and top reinforcing bars, respectively. Finally, the substitution
of these relations into Eq. (1) yields

M y1 + M y2 V b h c
V jh = A s1 f y1 + A s2 f y2 – ------------------------
- – ----------- (2)
h h

For simplicity, an overstrength factor on the beam moment

is not considered in this study.

Imposed deformations in joint regions

Here, imposed deformations in joint regions at three different
ductility levels are defined. To consider the required plastic
hinge deformation of a beam as an input parameter, the shear
strain of an interior beam-column joint is expressed in terms of
the required (or applied) strain εreq in the tension bars at the
beam-column interface for each ductility level.
Although the overstrength for beam moment is not considered,
the strain hardening of the reinforcing bars and the bond
stress along the yielded bars within the joint are used to
Fig. 2—Forces acting on interior beam-column joint. define εreq and the corresponding plastic deformation state

424 ACI Structural Journal/July-August 2011

 of both the beam plastic hinges and the joint region.
Figures 3(a) and (b) show the idealized strain-hardening
behavior and bond strength in the post-yield state developed
by Marti et al.13 The fictitious hardening modulus Esh and
fictitious bond strength fby in the post-yield state are
approximately 0.05Es and 0.2fb, respectively, where Es is
the elastic modulus, and fb is the elastic bond strength. Note
that the actual strain-hardening behavior is not linear and a
decrease in the bond strength may not occur at the specified
yield point; however, to simplify the relationship between
the plastic deformation of the beams and the strain of the
beam bars, these assumptions are made. In other words, the
post-yield parameters are introduced only to determine the
strain increment of the beam bars in the post-yield state rather
than to determine the change in the stress. Consequently, the
result is not overly sensitive to the post-yield parameters. In
this paper, the increase in joint shear due to the bar strain
hardening is neglected.
Figure 4(a) shows the equilibrium condition of the applied
forces Tb1 and Cb2 and bond stress u along the top beam bars
in the upper joint region, whereas Fig. 4(b) shows the strain
distributions of the top beam bar at different ductility levels.
Fig. 3—Behavior of steel reinforcing bars: (a) stress-strain The slopes of the strain along the bars within the joint can be
relationship; and (b) bond stress-bar strain relationship. calculated as follows

Fig. 4—(a) Equilibrium along beam bars passing through joint at three different ductility levels; (b) strain distributions; (c) stress
distributions; (d) bond stress distributions; and (e) admissible stress fields.

ACI Structural Journal/July-August 2011 425

 a c = ⎛ 0.25 + ----------------⎞ h c
dε s 4u N
------- = ----------- in the elastic state (3) (6)
dx db Es ⎝ h c b c fc′ ⎠

dε s 4f by where N is the column axial load, bc is the width of the

------- = ------------
dx d b E sh
- in the post-yield state (4) column, and fc′ is the specified concrete strength. Note that
the bond strength fb is not (or is just barely) reached inside
the “joint core” in State-I.
where u is the bond stress along the bar in the elastic state When the bond stress u along the beam bars reaches the
and has a value less than fb, εs is the bar strain, x is the bond strength fb at the outside of the joint core and the yield
distance from the beam-column interface on the tension side, penetration is not fully developed, the uniform stress fields
and db is the nominal bar diameter. When the bond stress u with continuous nodes along the beam and column bars
reaches fb, slippage between the concrete and the bars begins change to a fan-shaped stress field (refer to State-II versus
to occur. From the strain distributions, the bar stress and State-I in Fig. 4(e)). In State-II, the required strain magnitude
bond stress distributions are obtained, as shown in Fig. 4(c) εreq at the column face lies within the following two values
and (d). The bond stress along the column and beam bars
within the joint develops as a result of the truss mechanism. 4f by ⎛ E s d b⎞
Three admissible stress fields are possible. Figure 4(e) ε y + ------------ - < ε req ≤ ε y
- h c – a c – ε y ----------
E sh d b ⎝ 4f b ⎠
depicts different force-transfer mechanisms depending on (7)
the bond stress distribution along the beam bars. These three 4f by
- ( h – ac )
+ ------------
mechanisms represent the state of the elastic bond stress, the E sh d b c
state after reaching the bond strength, and the state after a
complete loss of bond strength.
In State-II (refer to Fig. 4(e)), the length ly of the yield
When the bond stress does not reach the bond strength, the shear penetration zone is determined as follows
force within a joint is transferred by arch and truss actions—that is,
by a diagonal strut with concentrated nodes (isolated nodal zones)
Es db d b E sh
and two uniform stress fields with continuous nodes along the h c – a c – ε y ----------
- < l y = ( ε req – ε y ) ------------
- ≤ hc – ac (8)
beam and column bars. In this stress condition (State-I hereafter), 4f b 4f by
the strain εreq of the beam bars at the column face is less than the
following value (refer to Fig. 4(e)) As the length of the yield penetration ly becomes longer,
the bond resistance along the beam bars becomes lower. The
4f by ⎛ E s d b⎞ bond stress in the tension-tension-compression node (T-T-C
ε req ≤ ε y + ------------
- h – a c – ε y ----------
- (5)
E sh d b ⎝ c 4f b ⎠ node; Fig. 5) decreases due to the full yield penetration in the
beam bars when the strain state satisfies the following state

where ac represents the compression zone depth in the column

4f by
above the joint, which may be estimated by the following ε y + ------------
- ( h – a c ) < ε req (9)
equation suggested by Paulay and Priestley12 E sh d b c

Fig. 5—(a) Stress fields of interior beam-column joint; and (b) strut-and-tie model.

426 ACI Structural Journal/July-August 2011

 In this condition (hereafter State-III; refer to Fig. 4(e)), the bars at the column face is illustrated in Fig. 5(a). This stress
length ly of the yield penetration zone is determined as follows field shows that both the primary diagonal strut D1 and the
adjacent diagonal struts D2 contribute to the diagonal strut
d b E sh mechanism, whereas the adjacent diagonal strut connecting
l y = ( ε req – ε y ) ------------
- > hc – ac (10) the T-T-C node N1 contributes to the truss mechanism.
4f by
Therefore, a strut-and-tie model based on these stress fields
can be constructed, as shown in Fig. 5(b). The internal force
Because the bond stress fields and the strain states in the flow patterns with the yield penetration of the beam bars
strut and fans interact with each other, both should be considered change the T-T-C node capacity.
to explain the strength degradation depending on the For a simple construction of a strut-and-tie model with
deformation. It is hypothesized that the deformation of a reasonable accuracy, the axis of compression force Cb2 at the
shear panel zone causes softening of the core concrete during top of the beam on the left-hand side of the joint is assumed
compression. The joint shear strain γJ is well represented by
to coincide with that of the tensile force Tb1 on the other side
the average tensile strain in the horizontal and vertical directions
of the joint, and the T-T-C node N1 is assumed to be located
within the joint (that is, (εh + εv)/2). First, the average
horizontal strain εh in the joint is estimated by the strain at the midpoint of the column depth. Note that the result is
distribution in the beam bars within the joint, and then the not overly sensitive to the location of N1. The horizontal
average vertical strain εv in the joint is estimated by the force shear force Vjh in the joint, which can be determined from
in the vertical reinforcement. Both are detailed in the Eq. (1), is transferred by three types of load paths. The first
following paragraphs. of these is caused by the primary diagonal strut D1
Figure 6 shows the horizontal strain distribution in the: (a) connecting two compression-compression-compression
top tension ties; (b) joint core; and (c) bottom tension ties. nodes (C-C-C nodes; refer to N2 in Fig. 5(b)) within the joint.
The average horizontal strain εh in the joint core can be The second is caused by the adjacent diagonal struts D2 and
calculated as the mean value of the strains in the top and horizontal ties Fh, which are activated by the confining effect
bottom ties. From the horizontal strain distributions illustrated of joint core reinforcement. The third path is attributable to a
in Fig. 6(b), the elongation of the top ties in the horizontal truss mechanism with the T-T-C node N1 via the vertical ties
direction is expressed as Fv, as mentioned in the previous section.
The strut angles of inclination of α1 and α2 are defined as
⎛1 1 1
⎜ --2- ε y l n1 + --2- ( ε y + εreq )ly – --2- ε c a c (State-I) h b′
⎜ tan α 1 = ------
- (12)
h c′
= ⎜ 1--- ⎛⎝ 2ε y – l n1 ----------b-⎞⎠ l n1 + 1--- ( ε y + εreq )ly – 1--- εc a c (State-II) (11)
4f
δ h, upp
⎜2 Es db 2 2
⎜
⎜ 1--- ⎛ 2ε – l -------------
4f by ⎞ 1
- l – --- ε a (State-III) h b′
⎝ 2 ⎝ req y E sh d b⎠ y 2 c c tan α 2 = ---------
- (13)
2h c′

where ln1 is the effective length of the T-T-C nodal zone as

(hc – ac – ly) (refer to Fig. 5(a)), and εc is the extreme fiber
compressive strain of the beam at the column face, which is
calculated using the strain compatibility and required tensile
strain εreq of the beam bar at the column face as (εreqab)/(d – ab).
Here, ab is the compressive stress block depth in the beam
and d is the effective beam depth. The term of (εcac/2) in
Eq. (11) is considered as a compressive axial deformation
component in the top ties.
Finally, the average strain εh,upp in the top ties is
calculated as δh,upp/hc) and the average strain εh,low in the
bottom ties is calculated in the same manner. The average
horizontal strain εh in the joint may be taken as the mean
value of the average of the strains in the top and bottom ties
(that is, (εh,upp + εh,low)/2).
On the other hand, the average vertical strain εv,avg is
simply approximated as Fv/EsAsv because it is within the
elastic deformation range, where Fv is the tensile force in
vertical reinforcement to form the truss mechanism via the
T-T-C node (N1; refer to Fig. 5(b)), and Asv is the area of
vertical reinforcement.

Strut-and-tie model
Using the relations identified in the preceding sections, a
strut-and-tie model is proposed for an interior beam-column
joint subject to moment reversals. One admissible stress Fig. 6—Stain distributions in horizontal direction: (a) top
field determined by the required strain level εreq in the beam chord; (b) joint core; and (c) bottom chord.

ACI Structural Journal/July-August 2011 427

where hc′ and hb′ are the distances between the main bars in effective nodal length ln1 in accordance with the states in
the columns and beams, respectively. Fig. 4(d). Hence, Vn1 is given by [n(πdb)ln1u], where n is the
Effective strengths of struts and tension ties—The number of beam bars in tension within the joint; db is the
behavior of concrete members is often governed by the nominal bar diameter; and for each state, the bond stress u
ultimate compressive strain of the concrete. It has been due to the shear force within Node N1 is expressed by
experimentally observed that the shear strength of joints with
typical details of transverse reinforcement and the
⎛
concrete bearing strength in the D-regions are mainly εy Es db
⎜ ---------------
- (State-I)
controlled by the softening behavior of cracked concrete ⎜ 4l n1
under compression.14-16 To estimate the effective strength u = ⎜ (15)
of a cracked diagonal strut, the equation proposed by ⎜ fb (State-II)
⎜
Vecchio and Collins17 is applied herein. Although the ⎝ 0 (State-III)
model by Vecchio and Collins17 was originally developed
from the unconfined panel tests, it might also be applied for
distressed beam-column joints in the nonlinear range (that The shear strength VT of the joint due to the truss action is
is, little lateral confinement at a relatively large drift ratio) limited either by the shear strength Vn1 of Node N1 or the strengths
reasonably well. The concrete softening model per Vecchio Fh_n and Fv_n of the horizontal and vertical ties as follows
and Collins17 is expressed as
⎛ V n1
f c′ ⎜
f ce - ≤ 0.85f c′
= --------------------------- (14) V T = minimum of ⎜ F h_n (16)
0.8 + 170ε 1 ⎜
⎝ F v_n cot α 1
where fce is the softened concrete strength, fc′ is the specified
concrete strength, and ε1 is the average principal tensile Note that the force in the vertical ties near Node N1 is mainly
strain of the cracked concrete. In this study, the direction of produced by the diagonal tensile stress, as it includes a very small
the diagonal strut is assumed to coincide with the direction contribution of the column bending moment to the tensile stress
of the principal compressive stress. The principal tensile in the bars. Also, it is very rare that the vertical reinforcement in
strain ε1 is obtained as εh + εv – ε2 by using Mohr’s circle, the middle part of the column reaches to the yield strain and thus,
where εh is the horizontal strain of (1/2)(εh,upp + εh,low) as in most cases, the strength of VT is governed by Vn1.
determined in the preceding section; εv is the vertical strain Capacity of Node N2—The nodal zone area of Node N2 is
of (Fv/EsAsv) as also determined in the preceding section; and defined as a region surrounded by the flexural compression
ε2 is the principal compressive strain, which is typically set equal blocks of the beam and column (with depths of ab and ac)
to –0.002 at the maximum strength of a diagonal strut. and the diagonal strut, as shown in Fig. 7. To estimate the
The strengths of horizontal tie Fh_n and vertical tie Fv_n capacity of Node N2, it is assumed that the load path within
are taken as the strengths of the joint hoops/crossties and the the joint is divided into a primary diagonal strut D1 and adjacent
column bars, respectively, in the joint core as Ash fyh and diagonal struts D2 generated by the horizontal tie. This adjacent
Asv fyv. Here, Ash is the cross-sectional area of joint hoops and diagonal strut D2 is a simplified representation of the stress
crossties in the direction of the column depth in a layer; Asv states represented by a fan-shaped stress field, as shown in
is the cross-sectional area of column vertical reinforcement at a Fig. 4(e). Also, the compression force due to the bond force
given point; and fyh and fyv are the specified yield strengths of the along the beam bars within Node N2 (that is, the horizontal
joint horizontal and column vertical reinforcement, respectively. shear force of Vn2 developed through Node N2) contributes to
Strength of Node N1—The strengths of Nodes N1 and N2 are the action of Strut D2.
defined, respectively. Based on the truss mechanism, the First, the strength CD1 of the primary diagonal strut D1 is
strength Vn1 to transfer the internal shear force developed limited by the effective strength of the cracked concrete strut as
through Node N1 is defined by the bond stress u and the
C D1 = a str1 bf ce (17)

where astr1 is the depth of the primary diagonal strut (refer to

Fig. 7), and b is the width of the strut.
Second, the strength CD2 of the adjacent diagonal strut D2
is determined as follows

⎛ ( a bf cos α 2 – V n2 ) ⁄ cos α 2
C D2 = minimum of ⎜ d2 ce (18)
⎝ ( F h – V n2 ) ⁄ cos α 2

where ad2 is the effective depth of Strut D2, and in that Strut D2
results from the fan-shaped (that is, triangular-shaped) stress
field in State-III (Fig. 4(e)), it is approximated as the average of
the depths (that is, the depth of the middle part of the fan)
Fig. 7—Geometries within C-C-C nodal zone (Node N2). as follows

428 ACI Structural Journal/July-August 2011

 1 whereas the third term refers to the shear strength due to the
a d2 = --- ( h b – 2a b ) cos α 2 (19) truss action. The strengths of the components of CD1, CD2,
2
and VT depend on the effective strengths of the struts, ties,
and nodes that were defined in the previous subsections.
where hb is the beam depth. As indicated in Eq. (18), the
Note that once the steel yielding occurs at the left end of ln1
strength Fh of the joint transverse reinforcement may govern
(Fig. 5(a)), the stress fields consist of a primary diagonal
the strut strength of CD2.
strut D1 and fan-shaped stress fields, as shown in Fig. 4(e),
In Eq. (18), Vn2 is the capacity to transfer the horizontal shear
and that the CD1 changes as the strain in the longitudinal bars
force developed through Node N2, which is defined as follows
increases, whereas the concrete strength in the fan-shaped
stress field is assumed to remain constant. As a result, the
V n2 = a str1 bf ce cos α 1 + a str2 b ( 0.85 f c′) cos α 2 + n ( πd b )l n2 f bc (20) primary strut crushes first as the steel strain increases.
In summary, the horizontal shear strength Vn_ jh of the
where astr2 is the depth of the adjacent diagonal strut, fbc is strut-and-tie model is dependent on the effective concrete
the bond stress within Node N2, and ln2 is the bonded length strength of the cracked diagonal strut fce, which then depends
within Node N2 (Fig. 7). To consider the beneficial effect of on the horizontal strain εh in the joint core. The value of εh is
confinement on the bond strength, it is assumed that the bond determined based on the horizontal strain distributions in the
strength along the bars within Node N2 is double the bond top and bottom tension bars (ties) and the bond strengths at
strength in the other tension ties (that is, fbc ≈ 2fb) based on different ductility levels.
the research by Orangun et al.18 and Muttoni et al.19
Setting the compressive stress contribution of Strut D2 Deformation of beam-column joint
equal to 0.85fc′ at the boundary of the node leads to the subassemblages
following equations. Based on the aforementioned strut-and-tie model, the
shear strength of an interior beam-column joint can be
n ( πd b )l n2 f bc + a str2 b ( 0.85f c′ ) cos α 2 = C D2 cos α 2 (21) related to the story drift ductility of the beam-column
element in terms of the tensile strain in the beam bars at the
a str1 sinα 1 + α str2 sin α 2 = a c – l n2 (22) column faces. The drift of the interior beam-column
subassemblage is influenced by the components due to the
elastic deformations of the beams and columns and by the
a str1 cos α 1 + α str 2 cos α 2 = a b (23) plastic hinge deformation of the beams and the joint shear
deformation. The drift of the subassemblage due to bond slip
within the joint is included as part of the joint shear deformation
Equations (22) and (23) are simple geometric relations
in the strut-and-tie model. For interior beam-column joints,
(Fig. 7). Solving the aforementioned equations gives dimensions
particularly for the joints with the same amount of top and
of astr1, astr2, and ln2 in terms of the strut strengths of CD2.
bottom beam bars, the bond slip in the plastic hinge deforma-
Shear strength of interior beam-column joints—Finally,
tion of beams is very small compared with that within the
the equilibrium condition within the upper chord in the
joint. Thus, the proposed strut-and-tie model may work
proposed strut-and-tie model results in the horizontal shear
better for these types of joints.
strength Vn_ jh as follows
Figure 8 illustrates the total drift ratio of the subassemblage as
a result of the deformation of these components. The interstory
V n_ jh = C D1 cos α 1 + C D2 cos α 2 + V T (24) drift ratio DR can be represented in terms of the deformation
of each component as follows
Here, the first and second terms on the right side of the
equation represent the shear forces transferred by the Δ C1 + Δ C2 ( Δ Be1 + Δ Be2 + Δ Bp1 + Δ Bp2 )
primary and adjacent diagonal strut actions with the - + -------------------------------------------------------------------- + γ J (25)
DR = -----------------------
confining effect of the joint reinforcement, respectively, hb lb

Fig. 8—Relationships of interstory drift and member deformations: (a) elastic deformation
of beam and column; (b) plastic hinge deformation; and (c) joint shear deformation.

ACI Structural Journal/July-August 2011 429

where ΔC1, ΔC2, ΔBe1, and ΔBe2 denote the interstory The interstory displacement of the subassemblage due to the
displacement contributions due to the elastic deformations of elastic deformation of the beams and columns can be calculated
the columns and beams, respectively; ΔBp1 and ΔBp2 are those using the effective stiffness of RC flexural members as
due to the plastic hinges of the right and left beams; hb and lb are suggested by ACI 318-08, Section 8.8,1 as follows
the hinge-to-hinge vertical and horizontal distances of the beam-
column subassemblages, respectively; and γJ is the shear strain Vc ( hi – hb ⁄ 2 )
3
of the joint, which is obtained using the average tensile Δ Ci = -----------------------------------
- (26)
strain (= (εh + εv)/2) in the horizontal and vertical directions in 3E col I col
the strut-and-tie model, as previously discussed. It should be
noted that the drift angle of γJ inherently includes the bond-slip 3
V b ( l bj – h c ⁄ 2 )
rotation at the beam-column interface, as the joint shear Δ Bej = ------------------------------------
- (27)
deformation is based on the average horizontal strain εh of the 3E Beam I Beam
joint that includes the contribution of the bond slip (that is,
integration of the bar strain) of the bars. In these equations, subscript i is 1 for the upper column
and 2 for the lower column; EcolIcol is the effective stiffness
Table 1—Dimensions and properties of test of the column; subscript j is 1 for the left beam and 2 for the
specimens right beam; EbeamIbeam is the effective stiffness of the beam;
and Vc and Vb are the shear forces acting on the columns and
hb1, hb2, tb, bb, hc, bc, N,
l1, m l2, m m the beams at the flexural yielding of the beam, respectively.
Specimen m mm mm mm mm kN
The interstory displacement due to the plastic hinges at the
B1 1.5 1.5 0.9 0.9 350 150 300 300 309
beams is determined based on the plastic strain εp and the
B2 1.5 1.5 0.9 0.9 350 300 300 300 301
plastic hinge length lp as follows
HH 1.5 1.5 0.9 0.9 350 200 300 300 353
Joh et al.15
HL 1.5 1.5 0.9 0.9 350 200 300 300 353
εp ⎞ ⎛ h l
LH 1.5 1.5 0.9 0.9 350 200 300 300 353 Δ Bpj = ⎛ ---------------
- l l – -----c – ---p⎞ (28)
MH 1.5 1.5 0.9 0.9 350 200 300 300 353
⎝ h b – a b⎠ p ⎝ bj 2 2 ⎠
A1 1.5 1 0.8 0.8 250 160 220 220 147
Fujii and
Morita9
A2 1.5 1 0.8 0.8 250 160 220 220 147 Also in this equation, subscript j is 1 for the left beam and
A3 1.5 1 0.8 0.8 250 160 220 220 441 2 for the right beam. The value of εp is defined as (εreq – εy),
A2 2 2 1.3 1.3 600 300 900 300 0 where εy is the yield strain of beam reinforcement. In this
Li et al.16 study, the plastic hinge length model developed by Paulay
M2 2 2 1.3 1.3 600 300 900 300 0
B12 2.4 2.4 1.7 1.7 610 356 457 457 325 and Priestley12 was used.
Beckingsale14
B13 2.4 2.4 1.7 1.7 610 356 457 457 2899
X1 1.8 1.8 1.2 1.2 500 250 450 300 0 COMPARISONS WITH TEST DATA
In the previous sections, a strut-and-tie model was
Xin20 X3 1.8 1.8 1.2 1.2 500 250 450 300 0
proposed that was capable of simulating the interior joint
X5 1.8 1.8 1.2 1.2 500 250 450 300 0
shear strength and its degradation. According to the
Notes: bb is beam width; 1 m = 3.28 ft; 1 mm = 0.03937 in.; 1 kN = 0.2248 kip.
proposed strut-and-tie model, the shear strength of the joint
is dependent on the bar deformation in adjacent beams and
Table 2—Material properties and reinforcement yield penetration into the joint, and the degraded shear
information strength for a given story drift ratio of beam-column frames
Joint hoops/ can be estimated. The approaches developed allow the
Concrete Beam reinforcement crossties
determination of the ductility limit or story drift capacity of
As1, fy1, As2, fy2, Ash , fyh, the system that is defined by joint shear failure. The verification
Specimen fc′, MPa mm2 MPa mm2 MPa mm2 MPa of the model is described in this section.
B1 21.3 398 371 398 371 226 307 Test specimens of interior beam-column joints in the literature
B2 20.8 398 371 398 371 226 307 were used to verify the capacity of the proposed strut-and-tie
HH 25.6 398 404 398 404 393 1165 model. A total of 16 specimens were used for the verification: six
Joh et al.15
HL 27.4 398 404 398 404 393 1165 tested by Joh et al.,15 three tested by Fujii and Morita,9 two tested
LH 26.9 398 404 398 404 226 377 by Li et al.,16 two tested by Beckingsale,14 and three tested by
MH 28.1 398 404 398 404 283 377 Xin.20 These specimens showed joint shear failure or
A1 40.2 628 409 628 409 170 291 degradation at various ductility levels.
Fujii and Table 1 shows the geometries of the specimens of the
A2 40.2 628 409 628 409 170 291
Morita9 interior beam-column joints for the calculation of their shear
A3 40.2 628 409 628 409 170 291
strengths and corresponding deformations. The properties of
A2 32.0 1900 463 982 460 603 499
Li et al.16 the beam and joint reinforcement and materials are described
M2 30.3 1900 463 982 460 603 499
in Table 2. All the specimens were designed to have a
B12 34.6 1710 298 1710 298 3545 336 column strength that was higher than the beam strength to
Beckingsale14
B13 31.4 1710 298 1710 298 2532 336 result in the plastic hinges at the beams.
X1 30.9 791 453 791 453 1459 348 In Fig. 9, the behavior of the beam-column subassemblages
20 X3 42.5 807 445 807 445 1459 348
Xin is presented in terms of the shear forces in the column Vc and
X5 60.7 942 492 942 492 1648 330 the interstory drift angle or interstory drift ratio DR. To
Notes: 1 MPa = 145 psi; 1 mm2 = 0.00155 in.2 estimate the strengths of the beam-column subassemblages in

430 ACI Structural Journal/July-August 2011

the horizontal direction, the magnitude of the flexural The response curves represent the elastic (and plastic)
deformation in the beams was first assumed and the joint behaviors of the joint specimens that experienced beam
shear strength was then assessed by the proposed strut-and- yielding and/or failed in a ductile manner. Once the response
tie model. In the range of moderate flexural deformations curve crosses the strength curve, the ductility of the beam-
in the beams, the estimated shear strengths of the joints are column subassemblage is limited by joint shear failure that
usually higher than the resulting shear forces due to the initiates at a drift ratio where two curves cross. In these ways,
flexural yielding of the beams because the shear failure of the ultimate capacity in terms of the story drift angle DRu
the joints does not control the system strength. The plastic and the corresponding column shear Vc_u are predicted by
behavior of the subassemblage may continue as long as the the proposed strut-and-tie model. In Table 3, comparisons
calculated joint shear strength is higher than the resulting between the experimental and analytical results show good
shear force upon the yielding of the beam bars. The agreement for most specimens with a wide range of drift
degradation of the joint shear strength, however, begins to capacities. Particularly, the deformation-based strut-and-tie
occur once the joint shear strength becomes lower than the model successfully predicted both the strengths and
resulting shear strength in the flexural failure mode. deformation capacities of the interior beam-column
Figure 9 shows the comparisons with test data, where the subassemblages that showed a substantial drop in strength
horizontal axis denotes the story drift angle or drift ratio DR of (for example, B1,15 A2,17 or M216), except for B1314 under
the selected beam-column assemblage, and the vertical axis significant axial loads (0.45Ag fc′, where Ag is the column
denotes the applied shear forces. The dotted lines are reproduced gross sectional area). Thus, further refinement may be
from the positive envelope of the hysteretic curves of the tested needed in cases where very high axial loads exist. Although
specimens and the solid lines emanating from the zero point Table 3 shows that the model substantially underestimates
(response curves) represent the behavior of the flexural failure the drift ratio at failure for Specimens HH,15 HL,15 and
mode. The latter uses an approximate bilinear relation with a MH,15 the model prediction is explicable given that the
strength of ((My1+My2)/(h) + (Vbhc)/(h)). The degrading solid measured decrease in strength was modest. For Specimens A1,9
lines (strength curves) represent the predicted behavior A2,9 and A3,9 the joint shear failure occurred without flexural
controlled by the joint shear strength according to the strut-and- yielding, as predicted. Based on the study presented in this paper,
tie model. the model could be further extended to investigate the ductility

Fig. 9—Comparisons with test results.

ACI Structural Journal/July-August 2011 431

 100 100
A1 20-
A2" 20 u;
z~ 80 "'
c.
;g_
z~ c.
g.
Flexural Behavior
.
.~
0

li
60 ·.Test Results
Flexural Behavior 15
. .
~
0

li
~
0

li
... Test Results
15
..!i
~

"
.<:
"
.<: "
.<: .<:
"'cE 40
10 "'
E
c "'cE 10
"'cE
~ ~ ~ ~
0 0 0 0
<J <J <J <J
20 5

0 0
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Drift Angle (rad) Drift Angle (rad)

120
A3 25
160
z~ 100 i;g, z~ 140
i;g,

.. .!i
20
~
.e"
0

li
Flexural Behavior
15
~
0
~
0 120

100
.
0
~
li
"
.c \ "
.c .c
"
.<:

"'E
c \ 10 "'E
c "'E
c
Flexural Behavior 80
"'cE
~ Test Results ~ 60
~
~
0 0 0
<J <J <J
Test Results · 40
Joint Shear Strength 5
20

0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 o.or 0.00 0.01 0.02 0.03 0.04 0.05 0.06
Drift Angle (rad) Drift Angle (rad)

800

160
812 160-
700
z~ 140
i;g, z~ 140
:g_
g.
600

."e
0

I'!
120

100
.."e .
0

li
0
"e
I'!
500 Joint Shear Strength
120

100
."=
~

.c .c .c 400 .c
"'cE Flexural Behavior 80
"'cE "'cE 300
80
"'cE
~ 60 ~ ~ 60 ~
0 0 0 0
<J <J <J 200 <J
·40 40

20 100 20

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Drift Angle (rad) Drift Angle (rad)

700 500

z~ 600
813
140-
.
c. z~
X1
100-
.
c.
400 ;g.
Joint Shear Strength 120 g. Joint Shear Strength .
."e
."e . "e
...lie
500 80
0 100 0 0 0
300
li 400 li
li 60
"
.c 80
"
.c "
.c Flexural Behavior-. .c
"'E
c 300
60 "'E
c "'E
c 200
Test Results:-\ \ "'E
c
~ TestR~s~l~~ ~
~
40
~
0 200 0 0 0
<J 40 <J <J <J
\
Flexural Behavior \ 20
100 20

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.00 0.01 0.02 0.03 0.04 0.05 0.06
Drift Angle (rad) Drift Angle (rad)

600

100
X'5 120 u;
z 500 c.
~
~
~ 100
~
.
.
"e
0

li
300
Joint Shear Strength

60
80 ~

.j .
~
0
~
0
Joint Shear Strength
80 ...e
0

li
"
.c
., ! .c
"'E
c 200
Te~~es~Hs~ c
80
"'E
c
~
40
~ ~ 40
~
0
<J 0 8 I 0
<J
100 Flexural Behavior 20 " Flexural Behavior j_
20

0 0 0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 O.Q1 0.02 0.03 0.04 0.05 0.06 0.07
Drift Angle (rad) Drift Angle (rad)

Fig. 9 (cont.)—Comparisons with test results.

432 ACI Structural Journal/July-August 2011

Table 3—Comparisons of strength and ACKNOWLEDGMENTS
deformation with test results The work presented in this paper was supported in part by the Korea
Science & Engineering Foundation (KOSEF) through the Korea Earthquake
Proposed Comparison, Engineering Research Center at Seoul National University, in part by
Test results model cal./test Interdisciplinary Research Grants (R01-2004-000-10290-0), and in part by the
Vc_u, DRu, Vc_u, DRu , (Vc_u)cal./ (DRu)cal./ U.S. DOT-RITA Grant (DTRT06-G-0016/OTCREOS9.1-27). The opinions,
findings, and conclusions in this paper are those of the authors and do not
Specimen kN rad kN rad (Vc_u)test (DRu)test
necessarily represent those of the sponsors.
B1 59.3 0.026 55.1 0.023 0.93 0.90
B2 63.6 0.022 57.1 0.025 0.90 1.11 REFERENCES
15
HH 63.8 0.031 59.1 0.042 0.93 1.34 1. ACI Committee 318, “Building Code Requirements for Structural
Joh et al. Concrete (ACI 318-08) and Commentary,” American Concrete Institute,
HL 66.8 0.032 60.0 0.042 0.90 1.32
Farmington Hills, MI, 2008, 473 pp.
LH 66.4 0.026 60.0 0.027 0.90 1.06
2. NZS 3101, “Concrete Structures Standard NZS 3101: Part 1,
MH 63.5 0.029 60.0 0.043 0.94 1.51 Commentary NZS 3101: Part 2,” Standards Association of New Zealand,
A1 66.6 0.013 71.6 0.015 1.08 1.13 Wellington, New Zealand, 1995, 520 pp.
3. Lin, C. M., and Restrepo, J. I., “Seismic Behaviour and Design of
Fujii and Morita9 A2 61.7 0.020 71.6 0.015 1.16 0.74
Reinforced Concrete Interior Beam-Column Joints,” Bulletin of the New
A3 67.8 0.014 71.7 0.020 1.06 1.40 Zealand Society for Earthquake Engineering, V. 35, No. 2, 2002, pp. 108-128.
A2 362.3 0.020 333.4 0.022 0.92 1.07 4. Wight, J. K., and Sozen, M. A., “Strength Decay of RC Columns
Li et al.16 under Shear Reversal,” Journal of the Structural Division, ASCE, V. 101,
M2 392.7 0.019 334.3 0.021 0.85 1.07
No. ST5, May 1975, pp. 1053-1065.
B12 197.7 0.034 179.4 0.045 0.91 1.32 5. ATC-6, “Seismic Design Guideline for Highway Bridge,” Applied
Beckingsale14
B13 195.2 0.037 210.5 — 1.08 — Technology Council, Redwood City, CA, 1981, 199 pp.
X1 152.9 0.030 139.3 0.035 0.91 1.16 6. Ang, B. G.; Priestley, M. J. N.; and Paulay, T., “Seismic Shear
Strength of Circular Reinforced Concrete Columns,” ACI Structural
Xin20 X3 164.1 0.040 142.6 0.044 0.87 1.10 Journal, V. 86, No. 1, Jan.-Feb. 1989, pp. 45-59.
X5 219.3 0.041 185.0 0.047 0.84 1.14 7. Hakuto, S.; Park, R.; and Tanaka, H., “Seismic Load Tests on Interior
Average — — — — 0.96 1.18 and Exterior Beam-Column Joints with Substandard Reinforcing Details,”
ACI Structural Journal, V. 97, No. 1, Jan.-Feb. 2000, pp. 11-25.
Standard deviation — — — — 0.09 0.20
8. Hakuto, S.; Park, R.; and Tanaka, H., “Effect of Deterioration of Bond
Note: 1 kN = 0.2248 kip. of Beam Bars Passing through Interior Beam-Column Joints on Flexural
Strength and Ductility,” ACI Structural Journal, V. 96, No. 5, Sept.-Oct.
and behavior of exterior or roof beam-column joints that are 1999, pp. 858-864.
9. Fujii, S., and Morita, S., “Comparison between Interior and Exterior
prone to shear failures or other shear-critical components. Beam-Column Joint Behavior,” Design of Beam-Column Joints for Seismic
Resistance, SP-123, J. O. Jirsa, ed., American Concrete Institute,
CONCLUSIONS Farmington Hills, MI, 1991, pp. 145-165.
10. Hwang, S.-J., and Lee, H.-J., “Analytical Model for Predicting Shear
The shear strength degradation of a joint due to the plastic Strength of Exterior Reinforced Concrete Beam-Column Joints for Seismic
deformation of the adjacent beam plastic hinge was investigated Resistance,” ACI Structural Journal, V. 96, No. 5, Sept.-Oct. 1999, pp. 846-858.
by using the redistribution of the internal forces due to the yield 11. Hwang, S.-J., and Lee, H.-J., “Model for Predicting Shear
penetration of the beam bars into the joint region. To explain Strength of Interior Reinforced Concrete Beam-Column Joints for
Seismic Resistance,” ACI Structural Journal, V. 97, No. 1, Jan.-Feb.
these phenomena, a strut-and-tie model for interior beam- 2000, pp. 35-44.
column joints was proposed. From this study, the following 12. Paulay, T., and Priestley, M. J. N., Seismic Design of Reinforced
conclusions are made: Concrete and Masonry Buildings, John Wiley & Sons, Inc., New York,
1. The proposed strut-and-tie model to account for the 1992, 744 pp.
13. Marti, P.; Alvarez, M.; Kaufmann, W.; and Sigrist, V., “Tension
plastic deformation of beam bars is a feasible solution for Chord Model for Structural Concrete,” Structural Engineering International,
predicting the degradation of shear strength and its degree International Association for Bridge and Structural Engineering, V. 8, No. 1,
for a given level of ductility (or story drift ratio). As the Nov. 1998, pp. 287-298.
behavior of most frames under earthquake loading mainly 14. Beckingsale, C. W., “Post-Elastic Behavior of Reinforced Concrete
Beam-Column Joints,” Report No. 80-20, Department of Civil Engineering,
depends on the flexural response of intended plastic hinge University of Canterbury, Christchurch, New Zealand, Aug. 1980, 379 pp.
regions and the stable flexural behavior is often terminated by 15. Joh, O.; Goto, Y.; and Shibata, T., “Influence of Transverse Joint and
the joint shear failure, the proposed deformation-based strut- Beam Reinforcement and Relocation of Plastic Hinge Region on Beam-
and-tie model can be applicable to the design of more Column Joint Stiffness Deterioration,” Design of Beam-Column Joints for
efficient and reliable seismic detailing to meet the Seismic Resistance, SP-123, J. O. Jirsa, ed., American Concrete Institute,
Farmington Hills, MI, 1991, pp. 187-224.
deformation requirements. 16. Li, B.; Wu, Y.; and Pan, T. C., “Seismic Behavior of Nonseismically
2. The proposed model can provide insight into the change Detailed Interior Beam-Wide Column Joints—Part I: Experimental Results
of the loading path within interior beam-column joints with and Observed Behavior,” ACI Structural Journal, V. 99, No. 6, Nov.-Dec.
2002, pp. 791-802.
consideration of the story drift ductility. This model can be 17. Vecchio, F. J., and Collins, M. P., “Modified Compression-Field
used as a rational means of assessing the strength of existing Theory for Reinforced Concrete Elements Subjected to Shear,” ACI
joints and also for new design, not only of high seismicity, Structural Journal, V. 83, No. 2, Mar.-Apr. 1986, pp. 219-231.
but also of low-to-moderate seismicity where a significant 18. Orangun, C. O.; Jirsa, J. O.; and Breen, J. E., “Reevaluation of Test
degree of ductile behavior is not required. Data on Development Length and Splices,” ACI JOURNAL, Proceedings
V. 74, No. 3, Mar. 1977, pp. 114-122.
3. The deformation-based strut-and-tie model may serve 19. Muttoni, A.; Schwartz, J.; and Thürlimann, B., Design of Concrete
as a rational tool for explaining the limited ductility behavior Structures with Stress Fields, Birkhäuser Verlag, Basel, Switzerland, 1997,
of RC members in shear. The bond strength degradation and 128 pp.
concrete softening behaviors were implemented in the proposed 20. Xin, X. Z., “Behaviour of Reinforced Concrete Interior Beam-
Column Joints Designed Using High Strength Concrete and Steel,” MS
strut-and-tie models, which consider the transitions of nodal thesis, Department of Civil Engineering, University of Canterbury,
zones with the introduction of appropriate strain fields. Christchurch, New Zealand, 1992, 121 pp.

ACI Structural Journal/July-August 2011 433