You are on page 1of 11

ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 106-S38

Drift Capacity of Reinforced Concrete Columns


by Sungjin Bae and Oguzhan Bayrak

This paper introduces simple closed-form equations to estimate the


lateral deformation capacity of reinforced concrete columns. The
proposed simple equations are derived from an idealized lateral
load-displacement behavior of a column and by taking the P-Δ
effect into account. The validity of the proposed equations is examined
using a large number of column tests. Estimated drift capacities
are compared with the experimentally-observed drift capacities.
The results show that the proposed equations provide conservative
lower-bound estimates of lateral drift capacities of columns that
satisfy the provisions of ACI 318-08, Chapter 21. It is also found
that for slender columns supporting high axial loads, even though
the detailing requirements of Chapter 21 are met, a satisfactory
lateral load performance can not be expected due to pronounced
secondary moments.

Keywords: curvature; displacement; drift; shear span-depth ratio; P-Δ effect.

INTRODUCTION
Predicting the deformation capacity of reinforced concrete
columns is of paramount importance in performance-based
earthquake engineering. For this purpose, the behavior of
concrete columns has been extensively investigated during
the last three decades. From these studies, analytical
procedures1-3 to model the column behavior have been
developed and performance-based concrete column design
expressions2,4-7 have been proposed. These analytical
procedures and performance-based design expressions,
however, have been verified using only a limited number of
column test results. As a result, there is a need to check the
validity of these analytical procedures and performance-based
design expressions using a more extensive column database.
To conduct a comprehensive study of column behavior,
Bae8 developed an analytical procedure that can be used to
estimate the moment-curvature and lateral load-displacement
relationships. Figure 1 presents a flowchart illustrating the Fig. 1—Flowchart of computer program.
main algorithm of the program. Phenomenological models
for concrete confinement and reinforcing bar buckling are
used in this analytical procedure. Flexural, bar slip, and shear
deformations are modeled and secondary displacements due Fig. 2. This figure indicates that the developed analytical
to the P-Δ effect are also calculated. In addition, Bae8 and procedure provides reasonably good drift capacity estimations
Bae and Bayrak12 identified the influence of axial load on for most cases. The error in the drift estimations is measured
the plastic hinge length and proposed a new expression for by using the ratio of the estimated drift capacity to the
plastic hinge length, as shown in Eq. (1). This new plastic experimental drift capacity. The mean and standard deviation
hinge length expression is included in this analytical procedure. are 1.03 and 0.29, respectively. The magnitude of the over-
A detailed discussion on this new plastic hinge length and under-estimations, however, can reach up to 50% in
expression can be found elsewhere.8,13 some cases. This indicates that even sophisticated analytical
procedures such as the one developed by Bae8 have certain
limitations in providing accurate and reliable deformation
l A
---p = 0.3 ⎛ -----⎞ + 3 ⎛ -----s⎞ – 0.1 ⎛ ---⎞ + 0.25 ≥ 0.25
P L capacity estimations consistently.
(1)
h ⎝ P o⎠ ⎝ A g⎠ ⎝ h⎠

ACI Structural Journal, V. 106, No. 4, July-August 2009.


MS No. S-2007-199 received May 27, 2007, and reviewed under Institute publication
The reliability of analytical procedures on the estimated policies. Copyright © 2009, American Concrete Institute. All rights reserved, including
deformation capacity was examined using a large number of 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 2010
column test results from a column database,14 as illustrated in ACI Structural Journal if the discussion is received by January 1, 2010.

ACI Structural Journal/July-August 2009 405


capacity) significantly, especially when large L/h coexist
ACI member Sungjin Bae is a Structural Engineer at Bechtel Corporation, Frederick,
MD. He received his BS and MS from Hanyang University, Seoul, Korea, and his PhD with high P/Po levels.
from the University of Texas at Austin, Austin, TX. He is a member of ACI Committee 209, With these motivations, simple closed-form equations to
Creep and Shrinkage of Concrete, and Joint ACI-ASCE Committee 441, Reinforced
Concrete Columns. His research interests include the behavior of concrete columns estimate the lateral deformation capacity of columns were
subjected to combined axial load and flexure, testing of large-scale reinforced derived and are reported herein. The derivation is based on
concrete columns, and performance-based design of concrete columns.
an idealized lateral load-displacement behavior of a concrete
ACI member Oguzhan Bayrak is an Associate Professor in the Department of Civil, column influenced by the P-Δ effect. The validity of the
Architectural, and Environmental Engineering and holds the Charles Elmer Rowe proposed simple equations in estimating deformation
Fellowship on Engineering at the University of Texas at Austin. He also serves as
the Director of the Phil M. Ferguson Structural Engineering Laboratory. He is a capacity is examined by using a large number of column test
member of ACI Committees 341, Earthquake-Resistant Concrete Bridges; E803, results from a database14 and the importance of the P-Δ
Faculty Network Coordinating Committee; and Joint ACI-ASCE Committees 441,
Reinforced Concrete Columns, and 445, Shear and Torsion.
effect on the column deformation capacity is discussed.

RESEARCH SIGNIFICANCE
The deformation capacity of concrete columns is an
important parameter in performance-based earthquake
engineering. This paper provides a simple tool that can be
used to estimate the deformation capacity of concrete columns
that satisfy the provisions of ACI 318-08, Chapter 21.20
Recently, several research studies7,15,19,21 have shown
that columns designed in accordance with provisions of
ACI 318-08, Chapter 21, may not display desired levels of
ductility under high axial loads. This issue is investigated by
using test results from a column database.14

LATERAL LOAD RESPONSE


OF CONCRETE COLUMNS
Moment-curvature and lateral load-displacement
Fig. 2—Comparison of estimated and experimental drift responses are commonly used to study column behavior. The
capacities. lateral load-displacement response of a column is affected by
the sectional behavior observed within the plastic hinge
region and the P-Δ effect. Accordingly, the loss of lateral
load resistance of a reinforced concrete column can be
caused by: 1) the degradation of moment capacity within the
plastic hinge region; and 2) the P-Δ effect, as illustrated in
Fig. 3. As shown in this figure, the lateral displacement
capacity of a column can be defined as the lateral displacement
at which a 20% loss in lateral load capacity occurs.7,19,22,23
It is important to note that the effect of axial load on column
behavior is two-fold. First, it influences the moment-curvature
response and hence affects the lateral load-displacement
response. Second, it produces the P-Δ effect as a column
deflects, which again affects the lateral load-displacement
response.
Figure 3 indicates that the lateral drift capacity of a reinforced
concrete column can be improved by reducing the degradation
Fig. 3—Moment-curvature and lateral load-displacement of moment resistance within the plastic hinge region and by
behavior of column. reducing the P-Δ effect. To minimize the degradation of
moment resistance, the amount and detailing of confining
Even though it is widely recognized that the P-Δ effect can reinforcement within the potential plastic hinge regions are
influence the column behavior, its effect on the deformation important. Based on this concept, many performance-based
capacity of concrete columns is not taken into account in design expressions2,4,6,7 have been proposed.
many performance-based column design expressions.2,4-7 The relationship between curvature and displacement was
As the P-Δ effect is a function of the shear span-depth ratio studied by Park and Paulay.24 They derived Eq. (2) for the tip
L/h and axial load P/Po, the investigation of the P-Δ effect displacement of a cantilever column using a plastic hinge
requires examining the L/h and P/Po simultaneously. In this concept (Fig. 4) and the second moment-area theorem. From
regard, even though the effect of the L/h and P/Po has been this equation, Park and Paulay24 further developed Eq. (3),
studied by many researchers,2,3,5-7,15-19 the influence of which relates the curvature ductility μφ to the displacement
these two parameters with respect to the P-Δ effect has not ductility μΔ.
been considered. Bae8 and Bae and Bayrak12 conducted full-
scale reinforced concrete column tests to study the P-Δ effect 2
on column deformation capacity. It was concluded that the φy L
- + ( φ – φ y )l p ( L – 0.5l p )
Δ t = Δ y + Δ p = ---------- (2)
P-Δ effect could affect the displacement capacity (or drift 3

406 ACI Structural Journal/July-August 2009


Table 1—Performance-based design methods
Minimum deformation capacity
Researchers Design expression for rectangular column sections for ductile behavior
A sh A ( φ u /φ y – 33ρ t m + 22 ) f c′ P
Watson et al.4 - = -----g ---------------------------------------------------
------ ----- --------------- – 0.006 μφ = 20
sb c Ac 111 f yh φf c′ A g

A sh f c, n f ce ′ ⎛ fy
P -⎞ + 0.13 ⎛ ρ ------
Wehbe et al.2 -------- = 0.1μ Δ ------
- 0.12 ------
- 0.5 + 1.25 ------------ - – 0.01⎞ μΔ = 10
st hc f ce′ f ye ⎝ f ce ′A g⎠ ⎝ t f s, n ⎠

⎧ P-⎞ 5 ⎫ ( μ φ )
1.15
Sheikh and Khoury6 A sh = α ⋅ ⎨ 1 + 13 ⎛ ---- ⎬ ----------------- A sh, ACI μφ = 20
⎝ P o⎠
⎩ ⎭ 29
f′ A 1 ----P-δ
Saatcioglu and Razvi7 ρ c = 14 ----c- -----g – 1 -------- δ = 2.5%
f yh A c k2 Po

l l
μ Δ = 1 + 3 ( μ φ – 1 ) ---p ⎛ 1 – 0.5 ---p⎞ (3)
L⎝ L⎠

where Δt is the tip displacement, Δy is the yield displacement,


Δp is the plastic displacement, φy is the yield curvature, φ is
the curvature, μΔ is the displacement ductility, and μφ is the
curvature ductility.
Equations (2) and (3) can be rewritten as follows

2 Fig. 4—Definition of plastic hinge length.24


φy L l l
- + ( φ – φ y ) ⎛ ---p⎞ ⎛ ---⎞ – 0.5 ⎛ ---p⎞ h
L 2
Δ t = ---------- (4)
3 ⎝ h ⎠ ⎝ h ⎠ ⎝ h⎠

⎛ l---p⎞ l p⎞
⎛ ---
⎝ h⎠ ⎝ h⎠
μ Δ = 1 + 3 ( μ φ – 1 ) --------- 1 – 0.5 --------- (5)
⎛L ---⎞ ⎛L ---⎞
⎝ h⎠ ⎝ h⎠

Equations (4) and (5) suggest that if the plastic hinge


length lp can be expressed in terms of the column depth h,
and if it is independent of the cantilever column height L, the
lateral displacement capacity (or lateral drift capacity) of a
column increases and the displacement ductility capacity
decreases as the L/h increases. This conclusion has been
supported by several researchers.7,17,18,25
It is worthy to note that the role of the P-Δ effect is not Fig. 5—Performance-based design expressions.
taken into account in deriving Eq. (2). Originally, Park and
Paulay24 derived this equation based on the concrete beam
behavior and later expanded its use to concrete columns by are used as design variables by Wehbe et al.2 and Saatcioglu
adjusting the plastic hinge length. Even with the adjustment and Razvi,7 respectively. Curvature ductility is also
of the plastic hinge length lp, however, the P-Δ effect on the frequently used by several other researchers.4,6 Proposed
relationship between different ductility parameters is not expressions for confinement reinforcement and corresponding
considered in Eq. (2). Therefore, the conclusions reached by minimum targeted deformation capacities are summarized in
examining Eq. (4) and (5) will only be valid when the Table 1. The required amount of confining reinforcement is
column behavior is not dominated by the P-Δ effect. calculated using a column with a square cross section of 24
x 24 in.2 (610 x 610 mm2), a concrete strength of 4300 psi
EXAMINATION OF PERFORMANCE-BASED (30 MPa), and Grade 60 reinforcing bars to study these
DESIGN EXPRESSIONS performance-based design methods. The minimum target
The relevant provision in Chapter 21 of ACI 318-0820 for deformation capacity is used for this purpose. A comparison
transverse reinforcement is intended to ensure that spalling of of the required amount of confining reinforcement is illustrated
cover concrete will not result in the loss of the axial load in Fig. 5. This figure shows large variations in the required
capacity of the columns. In this way, columns in plastic hinge amount of confining reinforcement among different design
regions are supposed to have adequate curvature capacity. methods. It is interesting to note that larger amounts of
The performance-based column design expressions,2,4,6,7 on the confining reinforcement are required when curvature ductility
other hand, are calibrated to achieve a certain deformation is used as a design parameter for high axial load levels.
capacity with respect to curvature ductility, displacement To examine the conservativeness of these performance-
ductility, or drift. Displacement ductility and drift capacity based column design expressions, results from 135 tests

ACI Structural Journal/July-August 2009 407


Table 2—Column database: summary
Section size, Concrete
in. x in. strength, Axial load Shear span-
Researchers (mm x mm) ksi (MPa) P/Po* depth ratio†
Azizinamini 18 x 18 5 to 6 0.2 to 0.4 3.0
et al.26 (457 x 457) (34 to 41)
24 x 24 4 to 6 0.2 to 0.5 5.0
8
(610 x 610) (28 to 41)
Bae
17.25 x 17.25 6
(438 x 438) (41) 0.5 7.0
Bayrak and 12 x 12 10 to 15 0.4 to 0.5 6.0
Sheikh22 (305 x 305) (69 to 103)
10 x 10 12
Galeota et al.27 (250 x 250) (83) 0.2 to 0.3 4.6

10 x 10 4
Kanda et al.28 (250 x 250) (28)
0.1 3.0

Légeron and 12 x 12 14 to 15 0.1 to 0.4 6.6


Paultre29 (305 x 305) (97 to 103)
Matamoros and 8x8 6 to 7
Sozen30 (203 x 203) (41 to 48) 0.0 to 0.3 3.0 Fig. 6—Column database: P-Δ effect.
15.7 x 15.7 4
Mo and Wang31 0.1 to 0.2 3.5
(400 x 400) (28) codes,43,44 a story drift capacity of 2.0 to 2.5% is expected
Muguruma 8x8 12 to 17 0.3 to 0.6 2.5
for special moment-resisting reinforced concrete frames
et al.32 (200 x 200) (83 to 117) designed for seismic effects. In this regard, a lateral drift
Ohno and 16 x 16 4 0.0 4.0
ratio of 2.5% is used as the deformation target. Figure 7
Nishioka33 (400 x 400) (28) shows that none of the performance-based design expressions
Paultre et al.34
12 x 12 11 to 16 0.4 to 0.6 6.6 guarantee a drift capacity of 2.5%. It is also interesting to note
(305 x 305) (76 to 110) that most performance-based design expressions, except
Saatcioglu and 14 x 14 5 0.2 to 0.4 4.7 Watson et al.,4 had problems when columns had large L/h
Grira35 (350 x 350) (34)
and support high P/Po. This is because the P-Δ effect on the
Saatcioglu and 14 x 14 5 to 6 column deformation capacity was ignored in these
0.0 to 0.1 2.9
Ozcebe16 (350 x 350) (34 to 41) performance-based column design expressions. The
Sakai et al.36
10 x 10 14 0.4 2.0 design expression developed by Watson et al.4 does not
(250 x 250) (97)
show the P-Δ effect problem on deformation capacity
Soesianawati 16 x 16 6 to 7 because it required a much larger amount of confining
0.1 to 0.3 4.0
et al.37 (400 x 400) (41 to 48)
reinforcement than the others for high axial load levels (as
9x9 17 shown in Fig. 5).
Sugano38 (225 x 225) (117)
0.4 to 0.7 2.0

22 x 22 5 0.1 to 0.3 3.0


Tanaka and (550 x 550) (34) ESTIMATION OF DEFORMATION CAPACITY:
Park39 16 x 16 4 SIMPLE METHOD
0.2 4.0
(400 x 400) (28) Figure 8 illustrates experimentally-obtained lateral load-
Thomson and 6x6 10 to 15 displacement relationships for concrete columns and the
0.0 to 0.2 3.9
Wallace40 (152 x 152) (69 to 103) nominal capacities in relation of the P-Δ effect. It can be
Watson and 16 x 16 6 observed that the slopes of descending branches of lateral
0.5 to 0.7 4.0
Park41 (400 x 400) (41) load-displacement relationships were in good agreement
24 x 15 4 with the slopes due to the P-Δ effect. The simple method to
Wehbe et al.2 (610 x 380) (28) 0.1 to 0.2 3.8
estimate the deformation capacity of concrete columns was
Xiao and 10 x 10 11 to 12 derived based on this observation of the influence of the P-Δ
0.1 to 0.2 2.0
Martirossyan42 (254 x 254) (76 to 83) effect on the lateral load-displacement relationship.
16 x 16 4 to 6
Zahn et al.3 (400 x 400) (28 to 41) 0.2 to 0.4 4.0
*
Lateral displacement capacity
Po = 0.85fc′(Ag− Ast) + fyl Ast.

To investigate the influence of the P-Δ effect on the lateral
Shear span-depth ratio is defined as ratio of distance between sections experiencing zero-
moment and maximum moment (L) to overall depth of column specimen (h). deformation capacity of a column in its simplest form, the
moment-curvature relationship of a reinforced concrete
conducted on reinforced concrete columns were used. These column was assumed to be elastic-perfectly-plastic. The
test results were obtained from a column database.14 Table 2 validity of this assumption for concrete columns that satisfy
shows the list of columns used for this study and some the ACI 318-0820 requirements will be discussed in the
important parameters, including P/Po and L/h. Figure 6 following section. With this assumption, the loss of lateral
illustrates the distribution of the column database with load resistance due to the degradation of moment capacity
respect to P/Po and L/h. Columns that were susceptible to was ignored and the resulting displacement capacity of a
significant P-Δ effect are marked in this figure. Column data column was directly attributed to the P-Δ effect.
are categorized with respect to L/h as well as P/Po to facilitate As can be seen in Fig. 3, the moment at the base of a
the comparison with respect to the P-Δ effect. Using this cantilever column was calculated as follows
column data, each performance-based design expression is
examined, as shown in Fig. 7. According to current building M=V·L+P·Δ (6)

408 ACI Structural Journal/July-August 2009


Fig. 7—Performance-based confining reinforcement design: corroboration with
experimental data.

Figure 9 illustrates the assumed moment-curvature relationship


and the corresponding lateral load-displacement relationship. It
is important to recognize that the post-peak part of the lateral
load-displacement relationship was solely controlled by the
P-Δ effect with this assumption. To simplify the derivation,
the y-intercept of the straight line extrapolated from the
descending branch of the lateral load-displacement relationship
was labeled as Vmax (Fig. 9). Alternatively, Vmax was envisioned
as the lateral load capacity of a very stiff column with a very
small yield displacement (Δy ≈ 0). As such, Vmax can be
defined as follows

Vmax = Mmax /L (7)

In a similar manner, the ultimate displacement Δult was


defined as the displacement where lateral load capacity
drops to zero due to the P-Δ effect.

M max
Δ ult = ------------ (8)
P

The slope of the descending branch of the assumed lateral load-


displacement relationship, shown in Fig. 9, was expressed as Fig. 8—Experimental lateral load-displacement relationships.

ACI Structural Journal/July-August 2009 409


Fig. 9—Influence of P-Δ effect on lateral load-displacement
curve.

Fig. 11—Drift capacity versus axial load level using Eq. (12)
(L/ h = 5).

Equation (11) indicates that the displacement capacity


(Δ80) due to the P-Δ effect is equal to 20% of the eccentricity e,
which is the ratio of the base moment capacity to the applied
axial load. The eccentricity of applied axial load acting on a
column was easily calculated from the P-M interaction
curve, which is illustrated in Fig. 10(a)

Fig. 10—Lateral displacement capacity: P-Δ effect. Drift capacity


Based on the displacement capacity presented in Eq. (11),
the drift capacity δ was calculated
1 = – ------------
k = -------- P (9)
Δ ult M max Δ 80 e
δ(%) = -------
- × 100 = 20 × --- (12)
L L
Based on the slope given in Eq. (9), the following expression
representing the post-peak part of the lateral load-displacement where L is the height of the cantilever column.
relationship was obtained Figure 11 presents the drift capacity of a typical concrete
column with an L/h of 5 supporting varying levels of axial
loads. Equation (12) was used in estimating the drift
⎛ ----------
V -⎞ = 1 + k ⋅ Δ = 1 – ------------
P ⋅Δ
(10) capacities shown in this figure. Figure 11 shows that the
⎝ Vmax ⎠ M max drift capacity of a column reduces as the axial load level
increases.
As discussed previously, the displacement capacity (Δ80)
was generally defined as the displacement at which a 20% Displacement ductility
drop in the lateral load capacity was observed. Because the The yield displacement was calculated as follows
actual maximum lateral load capacity (Vmax ′ in Fig. 9) was
also influenced by the P-Δ effect, the expression for the 20% 2
φy L
degradation of the lateral load capacity was somewhat Δ y = ----------
- (13)
3
complicated to include in a closed-form derivation. Typically,
the magnitude of the secondary moments at the maximum
′ ) is not very significant because
lateral load capacity (Vmax Using Eq. (11) and (13), the displacement ductility was
the magnitude of yield displacements is small. To simplify calculated as follows
the subsequent parts of the derivation, the 20% drop in the
lateral load capacity was calculated based on the maximum Δ 80 0.2e 0.6e
lateral load (Vmax) at zero displacement rather than the actual μ Δ = -------
- = ---------------- = ----------- (14)
Δy φ y L /3
2
φy L
2
maximum lateral load capacity (Vmax ′ ). As can be seen in
Fig. 9, this simplifying and conservative assumption yielded
underestimated displacement capacities (Δ80), especially As can be observed in Eq. (11), (12), and (14), the height
when high axial loads were acting on concrete columns. As of a cantilever column L influences the displacement
such, the displacement capacity (Δ80) was expressed as capacity, drift capacity, and the displacement ductility of a
column differently. The displacement capacity Δ80 is not
P affected by the height of a column. Here, it is important to
0.8 = 1 – ------------ ⋅ Δ 80 ; Δ 80 = 0.2 × e (11) note that the lateral displacement capacity of a column is
M max
defined as the lateral displacement at which a 20% loss in
lateral load capacity occurs and the loss of lateral load
where e = Mmax /P. resistance due to the degradation of moment capacity is not

410 ACI Structural Journal/July-August 2009


taken into account. On the other hand, the drift capacity δ
and displacement ductility μΔ are affected by the height of a
column. The drift capacity is inversely proportional to the
height of a cantilever column L, whereas the displacement
ductility is inversely proportional to the square of the height
of a cantilever column L2. Therefore, it can be concluded that
the displacement ductility μΔ is the most sensitive deformation
parameter to the column height among the three deformation
parameters (curvature ductility, displacement ductility, and
drift) and the curvature ductility, μφ , is the least sensitive to the
column height.

Proposed method to estimate deformation


capacity of columns
Simple closed-form equations that were used to estimate
the deformation capacity of reinforced concrete columns
were presented previously. A series of simplifying assumptions
were made in deriving Eq. (11), (12), and (14). For example,
the sectional behavior of a column was simplified to be
bilinear such that the degradation of the lateral load capacity Fig. 12—Column test results from Soesianawati et al.37
was solely related to the P-Δ effect. Because of this and other
simplifying assumptions, the deformation capacity estimations
have a certain level of curvature ductility in their moment-
provided by Eq. (11), (12), and (14) tend to be unrealistically
curvature relationships. To examine the validity of this
large for low axial load levels. To address this problem, the
assumption, the required curvature capacities for the use of
estimated drift capacities were capped at 4%. An examination
Eq. (15) are compared with the experimentally-obtained
of Fig. 11 indicates that this 4% drift capacity level
curvature capacities. Two column specimens, which were
corresponded to an axial load level of approximately 0.12Po
tested by Soesianawati,37 were used for this purpose. These
for columns with an L/h of 5. This upper limit placed on the
column specimens had square column sections and the
estimated drift capacities implies that an actual drift capacity
provided confining reinforcements were 33% and 55% of the
need not be calculated precisely for all practical purposes
ACI Code requirements, respectively. They were tested
when the drift capacity is 4%. In short, a drift capacity of 4%
under 0.1fc′ Ag and 0.3fc′ Ag.
is considered to be large enough for use in the performance-
Figure 12 shows the lateral load-displacement and
based design of reinforced concrete columns. With this upper
moment-curvature relationships of tested specimens. The
limit, the proposed method to estimate the deformation
displacement capacities estimated from Eq. (16) are marked
capacity of reinforced concrete columns can be summarized
in Fig. 12. The curvature values that correspond to the estimated
in Eq. (15) through (17)
displacement capacities are also illustrated in Fig. 12. The
moment-curvature relationships in Fig. 12 indicate that the
e
δ(%) = 20 × --- ≤ 4% (15) assumption of the bilinear moment-curvature relationship
L was valid and conservative for the curvature values required
by Eq. (16). It is important to note that the provided amount
Δ80 = 0.2 × e (16) of confining reinforcement for these two column specimens
was less than the amount required by ACI 318-08.20 Therefore,
0.6e it was concluded that if concrete columns are designed in
μ Δ = ----------2- (17) accordance with the provisions of ACI 318-08, Chapter 21,20
φy L the assumption of bilinear moment-curvature relationship
can be safely made in deriving Eq. (15) through (17). A more
where e = (Mn/P) ≤ 0.2 × L. detailed discussion of the estimated drift capacities for
It is important to note that the eccentricity (e = Mn/P) was columns conforming to the ACI 318-0820 requirements is
calculated from the P-M interaction diagrams plotted by provided in the subsequent sections where estimated drift
using the ACI 318 provisions for concrete stress block capacities are compared with the experimental values for
parameters. The aforementioned expressions indicate the 135 column tests.
calculated eccentricity shall be the smaller of Mn/P and 0.2 × L
to be consistent with the upper limit of drift capacity of 4%. Drift capacity estimation: simple and
By limiting the drift capacity up to 4%, Eq. (15) through (17) sophisticated methods
provided safe and lower-bound estimations of deformation In this section, the proposed simple method is compared
capacities for columns that met requirements of ACI 318-08, with a complicated computer-based analytical procedure
Chapter 21. (Fig. 1). For the purposes of comparison, a reinforced
concrete column with a 24 x 24 in.2 (610 x 610 mm2) square
Examination of assumed elasto-plastic moment- section and a longitudinal reinforcement ratio of 1.0% was
curvature relationship used (Fig. 13). Two different cantilever column heights of
A bilinear moment-curvature relationship was assumed in 72 and 120 in. (1830 and 3050 mm) was used, which
the derivation of Eq. (15) through (17). To obtain conservative resulted in L/h values of 3 and 5, respectively. The confining
estimations of deformation capacities from Eq. (15) through reinforcement was provided in accordance with the requirements
(17), this bilinear assumption requires concrete columns to of Chapter 21 of ACI 318-08.20

ACI Structural Journal/July-August 2009 411


strain hardening of longitudinal steel can contribute to the
conservativeness of estimations.

Drift capacity estimation: comparison with


column database
To investigate the validity of Eq. (15), the drift capacities
obtained through the use of Eq. (15) were compared with the
experimental drift capacities. Results from 135 column tests
were obtained from the column database14 and used for this
purpose. The results are shown in Fig. 14. This figure
illustrates that there was a strong correlation between the
lower-bound of experimentally-obtained drift capacities and
Fig. 13—Comparison of predicted drift capacities. the drift capacities estimated by using the proposed method.
The proposed method provides safe estimations for drift
capacity in most cases. As can be seen in Fig. 14, some test
data fall short of the drift capacities estimated by using Eq. (15).
Most of those data, however, were from columns that do not
meet the seismic design and detailing provisions (Chapter 21) of
the ACI Code.20
To study the conservativeness of the drift capacity estimations
and to evaluate the performance of columns that satisfy
Chapter 21 provisions of ACI 318-08, test data were categorized
into two groups: 1) columns that met or exceeded the
confinement reinforcement provisions of Chapter 21; and
2) columns that contained less confining reinforcement than
that required by the seismic design provisions of ACI 318-08.20
The results for columns that satisfy the ACI Code20
requirements are shown in Fig. 15(a). The results presented in
Fig. 15(a) indicate that Eq. (15) traces the lower bound of the
data consistently for all data, except for a few columns tested
by Sugano.38 The drift capacities of some of the column
specimens tested by Sugano38 were slightly overestimated
by the proposed method. The columns tested by Sugano38
were made from very high-strength concrete ( fc′ = 120 MPa
[17,404.5 psi]) and had a small L/h of 2. Columns with such
small L/h can be affected by a shear failure mechanism. A
shear failure can limit the drift capacity of a column due to
its brittle nature. In addition to the small L/h, the brittleness
of the very high-strength concrete used may also have been
a contributing factor to the limited drift capacity. In short, the
proposed method (Eq. (15)) resulted in safe estimations of
drift capacity for all columns that satisfy the provisions of
Fig. 14—Equation (15) and test results. ACI 318-08, Chapter 21,20 except some of the columns tested
by Sugano.38 It is also important to note that the upper lateral
drift limit of 4% provided a safe lower-bound estimation for
The results shown in Fig. 13 illustrate that the drift capacities columns tested under low axial load levels.
estimated by Eq. (15) were close to the drift capacities Researchers2,4,6,19,21,45 have reported that the ACI 318-0820
obtained from a more rigorous analysis where numerous requirements may be unsafe for columns at high axial load
material nonlinearities were considered. It is interesting to levels and proposed more strict requirements for confining
note the use of simplified expressions yielded lower drift reinforcement at high axial load levels, as can be seen in Fig. 5.
capacities than those obtained from the rigorous analysis. The seismic detailing and design provisions of most of the
The differences between the drift capacity estimations current design codes are aimed at achieving a story drift ratio
obtained through the use of these methods increased with of 2.0 to 2.5% for moment-resisting reinforced concrete
increasing L/h. frames.43,44 Therefore, concrete columns with drift capacities
The conservative estimations obtained by using Eq. (15) that are larger than 2.5% can be considered as ductile
were attributed to the following: 1) the use of the ideal columns. Based on this criterion, Fig. 15(a) shows that all
maximum lateral load Vmax without the P-Δ effect instead of columns that satisfy the ACI 318-0820 requirements had a
the obtainable maximum lateral load capacity Vmax ′ of the minimum drift capacity of 2.5% when their L/h ranged from
lateral load-displacement relationship in deriving the equations 2 to 5 and the axial load was less than 0.52Po, which is the
(Fig. 9); and 2) the use of the nominal moment capacity Mn maximum permissible axial load as per ACI 318-08.20 Only
using ACI 318-0820 provisions instead of the actual columns supporting axial loads higher than 0.52Po and with
maximum moment capacity Mmax in constructing the very small L/h (less than 2) or very large L/h (greater than 5)
bilinear moment-curvature relationships. Ignorance of the may not achieve a drift capacity of 2.5% because of brittle
possible strength increase due to concrete confinement and shear failures for L/h < 2 or very large P-Δ moments. As

412 ACI Structural Journal/July-August 2009


such, it can be concluded that columns that follow the ACI
318-0820 requirements will be ductile as long as reasonable
L/h and P/Po levels are used. A detailed discussion on the
relationship between L/h and P/Po is provided in the
following section.
Figure 15(b) illustrates the drift capacity estimations for
columns that contained less lateral reinforcement than that
required by the provisions of ACI 318-08, Chapter 21.20
Once again, Eq. (15) provides a reasonable lower bound to
the experimental drift capacities in most cases. The percentage
of overestimated drift capacities, however, was considerably
higher for this group of columns. More specifically, the
proposed method could not provide conservative drift
estimations for most columns tested by Sakai36 and
Thomson and Wallace,40 and some of the columns tested by
Galeota et al.27 These specimens were all high-strength concrete
columns (fc′ = 70 to 100 MPa [10,152.6 to 14,503.7 psi]). For
the specimens tested by Galeota27 (L/h = 4.6), Eq. (15)
provided safe estimations for the drift capacities for all of
the columns that contained more than 30% of the lateral
reinforcement required by the provisions of ACI 318-08,
Chapter 21.20 The drift capacities of most of the columns
tested by Sakai36 (L/h = 2) and Thomson and Wallace40
(L/h = 3.9) were overestimated. Such lightly-confined
concrete columns with small or moderate L/h were
vulnerable to experiencing large shear deformations and,
hence, experienced shear failure at low drift levels.
In summary, the proposed simple equations can provide
conservative lower-bound estimates of deformation capacity
for concrete columns that satisfy the ACI 318-0820 requirements.
It can be also concluded that the drift capacity decreases as the
axial load increases and the L/h increases due to the P-Δ effect.
It is also interesting to note that large drift capacities can be
obtained at low axial loads regardless of the L/h of columns
because the P-Δ effect is negligibly small for low axial loads.

Design recommendations
It was previously discussed that the P-Δ effect influenced
the lateral deformation capacity of concrete columns. The P-Δ
effect was more pronounced for high axial loads and large L/h.
Therefore, to ensure the ductile behavior of concrete
columns, it is important to control the P-Δ effect by limiting
axial load levels for given a L/h.
It was also discussed that a simple expression (δ = 20 × e/L ≤
4%) provides conservative estimations of drift capacity for
concrete columns that are designed in accordance with the
Chapter 21 requirements of the ACI 318 Code.20 Using this
simple expression, the drift capacity of columns was evaluated
for the target drift capacity of 2.5%.
Figure 16 illustrates that for a given target drift capacity
(for example, 2.5%), the permissible axial load limit
decreases as the L/h of a column increases. The permissible
P/Po to ensure a drift capacity of 2.5% is summarized in
Table 3. Therefore, it can be concluded that to ensure stable
lateral deformation capacity of concrete columns, it is
important to satisfy not only the requirements of the Chapter
21 provisions of ACI 318-08,20 but also the L/h presented in Fig. 15—Equation (15): Columns that meet Chapter 21
Table 3. It is also important to note that to achieve satisfactory provisions versus lightly-confined columns.
drift capacities, the use of relatively smaller L/h and lower P/Po
are recommended. As large shear forces are developed in Table 3 indicates that sufficient deformation capacity (drift
columns with very small L/h, however, the column shear capacity of at least 2.5%) will be guaranteed when reinforced
strength has to be checked for very small L/h. L/h values of concrete columns are designed in accordance with the
most building columns in high seismic areas range between Chapter 21 provisions of ACI 318-0820 with specified limits
2 and 4 (or the story height-to-column depth ratio of 4 to 8). of L/h and P/Po.

ACI Structural Journal/July-August 2009 413


columns, it is recommended to limit the levels of axial
load at large L/h (Fig. 16). It is important to recognize
that this study has primarily focused on the concrete
columns whose behavior is governed by flexure.
Because shear failure is not considered in the analysis
methods employed to estimate the deformation
capacity, the shear strength of columns with very small
L/h has to be considered explicitly in design.

NOTATION
Ag = cross-sectional area of column
As = cross-sectional area of longitudinal reinforcement
Ash = total cross-sectional area of confining reinforcement
Ash,ACI = total cross-sectional area of confining reinforcement required
by Chapter 21 of ACI 318-0820
e = M/P = eccentricity of axial load
fc′ = compressive strength of concrete
fy = yield strength of reinforcement
h = overall depth of column
L = height of cantilever column
lp = plastic hinge length
Mmax = maximum moment capacity of column
Mn = nominal moment capacity as per ACI 318-0820
Fig. 16—Interaction among drift capacity, axial load, and P = axial load applied to column
shear span-depth ratio. Po = 0.85fc′ (Ag – Ast) + fy Ast = nominal axial load capacity as per
ACI 318-0820
V = lateral load acting on column
Table 3—Recommended axial load levels for Δ80 = displacement at 20% reduction of lateral load capacity
various shear span-depth ratios* Δp = plastic displacement
Δt = tip displacement
Shear span-depth ratio (L/h)† Permissible axial load level (P/Po)
Δy = yield displacement
2 to 3 0.4 δ = Δt80/L = drift
3 to 4 0.3 φ = curvature
φe = elastic curvature
4 to 5 0.2
φp = plastic curvature
5 to 6 0.15 φu = maximum curvature at column
*Minimum drift capacity = 2.5%. φy = yield curvature
†Refer to Fig. 16 for definition of L/h. μΔ = Δ/Δy = displacement ductility
μφ = φ/φy = curvature ductility
ρl = ratio of longitudinal reinforcement to gross sectional area
SUMMARY AND CONCLUSIONS
A simple method that can be used to estimate the lateral REFERENCES
deformation capacity of a reinforced concrete column is 1. Yalcin, C., and Saatcioglu, M., “Inelastic Analysis of Reinforced
presented. The validity of the proposed simple method is Concrete Columns,” Computers and Structures, V. 77, No. 5, July 2000,
examined using results from a large number of column pp. 539-555.
2. Wehbe, N. I.; Saiidi, M. S.; and Sanders, D. H., “Effects of Confinement
tests obtained from a column database.14 Based on the and Flares on the Seismic Performance of Reinforced Concrete Bridge
research summarized in this paper, the following conclusions Columns,” Report No. CCEER-97-2, Engineering Research and Development
can be made: Center, University of Nevada, Reno, NV, Sept. 1997, 399 pp.
• A simple closed-form expression that can be used to 3. Zahn, F. A.; Park, R.; and Priestley, M. J. N., “Design of Reinforced
Concrete Bridge Columns for Strength and Ductility,” Research Report 86-7,
estimate the drift capacity of reinforced concrete Department of Civil Engineering, University of Canterbury, Christchurch,
columns was presented. Through the use of this simple New Zealand, Mar. 1986, 330 pp.
expression, the drift capacities of columns satisfying 4. Watson, S.; Zahn, F. A.; and Park, R., “Confining Reinforcement for
Chapter 21 provisions of ACI 318-0820 were estimated Concrete Columns,” Journal of Structural Engineering, ASCE, V. 120,
in a conservative manner when their L/h values were No. 6, June 1994, pp. 1798-1824.
5. Watson, S., and Park, R., “Simulated Seismic Load Tests on Reinforced
larger than 2. As such, this expression is recommended Concrete Columns,” Journal of Structural Engineering, ASCE, V. 120,
for use in performance-based earthquake engineering; No. 6, June 1994, pp. 1825-1849.
• The examination of the proposed method with 135 6. Sheikh, S. A., and Khoury, S. S., “Performance-Based Approach for
column test results indicated that the drift capacity of the Design of Confining Steel in Tied Columns,” ACI Structural Journal, V. 94,
No. 4, July-Aug. 1997, pp. 421-431.
concrete columns was significantly influenced by the 7. Saatcioglu, M., and Razvi, S. R., “Displacement-Based Design of
P-Δ effect as well as the sectional behavior. As a result, Reinforced Concrete Columns for Confinement,” ACI Structural Journal,
the drift capacity decreased as the axial load increased V. 99, No. 1, Jan.-Feb. 2002, pp. 3-11.
and the L/h increased. It is also interesting to note that 8. Bae, S., “Seismic Performance of Full-Scale Reinforced Concrete
large drift capacities were obtained at low axial loads Columns,” PhD dissertation, University of Texas at Austin, Austin, TX,
Dec. 2005, 311 pp.
regardless of the L/h of the columns because the P-Δ 9. Bae, S,; Mieses, A. M.; and Bayrak, O.,“Inelastic Buckling of Reinforcing
effect was negligibly small at low axial loads; and Bars,” Journal of Structural Engineering, ASCE, V. 131, No. 2, Feb. 2005,
• Based on the research reported in this paper, design pp. 314-321.
recommendations are made to slightly improve the 10. Alsiwat, J. M., and Saatcioglu, M.,“Reinforcement Anchorage Slip
under Monotonic Loading,” Journal of Structural Engineering, ASCE,
Chapter 21 provisions or commentary of ACI 318-0820 V. 118, No. 9, Sept. 1992, pp. 2421-2438.
that were found to be satisfactory for most cases. To 11. Lehman, D. E., and Moehle, J. P.,“Seismic Performance of Well-
guarantee sufficient deformation capacity of concrete Confined Concrete Bridge Columns,” PEER Report 1998-01, College of

414 ACI Structural Journal/July-August 2009


Engineering, University of California-Berkeley, Berkeley, CA, Dec. 2000, Journal, V. 97, No. 4, July-Aug. 2000, pp. 591-601.
286 pp. 30. Matamoros, A. B., and Sözen, M. A., “Drift Limits of High-Strength
12. Bae, S., and Bayrak, O., “Seismic Performance of Full-Scale Reinforced Concrete Columns Subjected to Load Reversals,” Journal of Structural
Concrete Columns,” ACI Structural Journal, V. 105, No. 2, Mar.-Apr. 2008, Engineering, ASCE, V. 129, No. 3, Mar. 2003, pp. 297-313.
pp. 123-133. 31. Mo, Y. L., and Wang, S. J., “Seismic Behavior of RC Columns with
13. Bae, S., and Bayrak, O., “Plastic Hinge Length of Reinforced Various Tie Configurations,” Journal of Structural Engineering, ASCE,
Concrete Columns,” ACI Structural Journal, V. 105, No. 3, May-June V. 126, No. 10, Oct. 2000, pp. 1122-1130.
2008, pp. 290-300. 32. Muguruma, H.; Watanabe, F.; and Komuro, T., “Applicability of
14. Structural Performance Database, University of Washington, High Strength Concrete to Reinforced Concrete Ductile Column,”
www.ce.washington.edu/~peera1/. Transactions of the Japan Concrete Institute, V. 11, 1989, pp. 309-316.
15. Bayrak, O., “Seismic Performance of Rectilinearly Confined High 33. Ohno, T., and Nishioka, T., “An Experimental Study on Energy
Strength Concrete Columns,” PhD dissertation, University of Toronto, Absorption Capacity of Columns in Reinforced Concrete Structures,”
Toronto, ON, Canada, 1999, 339 pp. Proceedings of the JSCE, V. 1, No. 2, Oct. 1984, pp. 137-147.
16. Saatcioglu, M., and Ozcebe, G., “Response of Reinforced Concrete 34. Paultre, P.; Légeron, F.; and Mongeau, D., “Influence of Concrete
Columns to Simulated Seismic Loading,” ACI Structural Journal, V. 86, Strength and Transverse Reinforcement Yield Strength on Behavior of
No. 1, Jan.-Feb. 1989, pp. 3-12. High-Strength Concrete Columns,” ACI Structural Journal, V. 98, No. 4,
17. Pujol, S.; Sözen, M.; and Ramírez, J., “Transverse Reinforcement for July-Aug. 2001, pp. 490-501.
Columns of RC Frames to Resist Earthquakes,” Journal of Structural 35. Saatcioglu, M., and Grira, M., “Confinement of Reinforced Concrete
Engineering, ASCE, V. 126, No. 4, Apr. 2000, pp. 461-466. Columns with Welded Reinforcement Grids,” ACI Structural Journal, V. 96,
18. Kowalsky, M. J., “Deformation Limit States for Circular Reinforced No. 1, Jan.-Feb. 1999, pp. 29-39.
Concrete Bridge Columns,” Journal of Structural Engineering, ASCE, 36. Sakai, Y.; Hibi, J.; Otani, S.; and Aoyama, H., “Experimental Study
V. 126, No. 8, Aug. 2000, pp. 869-878. on Flexural Behavior of Reinforced Concrete Columns Using High-
19. Sheikh, S. A., and Khoury, S. S., “Confined Concrete Columns with Strength Concrete,” Transactions of the Japan Concrete Institute, V. 12,
Stubs,” ACI Structural Journal, V. 90, No. 4, July-Aug. 1993, pp. 414-431. 1990, pp. 323-330.
20. ACI Committee 318, “Building Code Requirements for Reinforced 37. Soesianawati, M. T.; Park, R.; and Priestley, M. J. N., “Limited
Concrete (ACI 318-08) and Commentary,” American Concrete Institute, Ductility Design of Reinforced Concrete Columns,” Research Report 86-10,
Farmington Hills, MI, 2008, 423 pp. Department of Civil Engineering, University of Canterbury, Christchurch,
21. Sheikh, S. A.; Shah, D. V.; and Khoury, S. S., “Confinement of High- New Zealand, Mar. 1986, 208 pp.
Strength Concrete Columns,” ACI Structural Journal, V. 91, No. 1, 38. Sugano, S., “Seismic Behavior of Reinforced Concrete Columns
Jan.-Feb. 1994, pp. 100-111. which used Ultra-High-Strength Concrete,” Paper No. 1383, Eleventh
22. Bayrak, O., and Sheikh, S. A., “High-Strength Concrete Columns World Conference on Earthquake Engineering, Acapulco, Mexico, 1996.
under Simulated Earthquake Loading,” ACI Structural Journal, V. 94, No. 6, 39. Tanaka, H., and Park, R., “Effect of Lateral Confining Reinforcement
Nov.-Dec. 1997, pp. 708-722. on the Ductile Behavior of Reinforced Concrete Columns,” Research Report
23. Légeron, F., and Paultre, P., “Behavior of High-Strength Concrete 90-2, Department of Civil Engineering, University of Canterbury,
Columns under Cyclic Flexure and Constant Axial Load,” ACI Structural Christchurch, New Zealand, June 1990, 458 pp.
Journal, V. 97, No. 4, July-Aug. 2000, pp. 591-601. 40. Thomson, J., and Wallace, J., “Lateral Load Behavior of Reinforced
24. Park, R., and Paulay, T., Reinforced Concrete Structures, John Wiley Concrete Columns Constructed Using High-Strength Materials,” ACI
and Sons, New York, 1975, 769 pp. Structural Journal, V. 91, No. 5, Sept.-Oct. 1994, pp. 605-615.
25. Berry, M. P., and Eberhard, M. O., “Practical Performance Model for 41. Watson, S., and Park, R., “Design of Reinforced Concrete Frames of
Bar Buckling,” Journal of Structural Engineering, ASCE, V. 131, No. 7, Limited Ductility,” Research Report 89-04, Department of Civil Engineering,
July 2005, pp. 1060-1070. University of Canterbury, Christchurch, New Zealand, Jan. 1989, 232 pp.
26. Azizinamini, A.; Corley, W. G.; and Johal, L. S., “Effects of Transverse 42. Xiao, Y., and Martirossyan, A.,“Seismic Performance of High-
Reinforcement on Seismic Performance of Columns,” ACI Structural Strength Concrete Columns,” Journal of Structural Engineering, ASCE,
Journal, V. 89, No. 4, July-Aug. 1992, pp. 442-450. Mar. 1998, pp. 241-251.
27. Galeota, D.; Giammatteo, M. M.; and Marino, R., “Seismic Resistance 43. International Code Council (ICC), “International Building Code,”
of High Strength Concrete Columns,” Paper No. 1390, Eleventh World Falls Church, VA, 2006, 663 pp.
Conference on Earthquake Engineering, Acapulco, Mexico, 1996. 44. International Conference of Building Officials (ICBO), “Uniform
28. Kanda, M.; Shirai, N.; Adachi, H.; and Sato, T., “Analytical Study on Building Code, Volume 2—Structural Engineering Design Provisions,”
Elasto-Plastic Hysteretic Behaviors of Reinforced Concrete Members,” Whittier, CA, 1997, 492 pp.
Transactions of the Japan Concrete Institute, V. 10, 1988, pp. 257-264. 45. Li, B., and Park, R., “Confining Reinforcement for High-Strength
29. Légeron, F., and Paultre, P., “Behavior of High-Strength Concrete Concrete Columns,” ACI Structural Journal, V. 101, No. 3, May-June
Columns under Cyclic Flexure and Constant Axial Load,” ACI Structural 2004, pp. 314-324.

ACI Structural Journal/July-August 2009 415

You might also like