A Reliability-Based Deformation Capacity Model for ACI 318
Compliant Special Structural Walls
Abstract
The underlying premise of the ASCE 7 and ACI 318
provisions is that special structural walls satisfying the
provisions of ACI 318-14 §18.10.6.2 through §18.10.6.4
possess adequate drift capacity to exceed the expected drift
demand determined from ASCE 7-10 analysis procedures.
However, observations from laboratory tests and strong
earthquakes, e.g., the 2010 Chile earthquake and the 2010 E-
Defense tests, where significant damage was noted at wall
boundaries due to concrete crushing, rebar buckling, and
lateral instability, have raised concerns that current design
provisions are inadequate, especially for “thin” walls,
although it is not clear what attributes should be used to
define a “thin” wall. Recent studies have identified that
deformation capacity of code compliant walls is primarily a
function of wall cross-section geometry, neutral axis depth,
wall shear stress demands, and configuration of boundary
transverse reinforcement (i.e., use of overlapping hoops
versus a perimeter hoop with intermediate legs of crossties);
however, these parameters are not adequately considered in
ACI 318-14. To address this issue, this study proposes a new
reliability-based design requirement where a drift demand-to-
capacity ratio (DDCR) check is performed to provide a low
probability that roof drift demands exceed roof drift capacity
at strength loss (e.g., 10% probability of lateral strength loss
under Design Earthquake). In general, slender wall cross
( )
sections lw b > 15 with large neutral axis depth relative to
width of flexural compression zone ( c b > 3) and shear
stresses approaching the ACI 318-14 limit (10 f 'c ) tend
to be screened out for redesign to reduce the probability of
collapse under MCE level shaking.
Saman A. Abdullah, PhD Candidate
University of California, Los Angeles
Los Angeles, CA
John W. Wallace, Professor
University of California, Los Angeles
Los Angeles, CA
Introduction
In the last version of ACI 318 in 2014, extensive revisions
were introduced to require more stringent detailing
( )
requirements for thin, slender hw lw ≥ 2.0 walls, include a
limit on wall slenderness (h
u )
b ≤ 16 , require a minimum
width of flexural compression zone (b > 12 in.) for sections
( )
that are not tension-controlled c lw ≥ 3 8 , and require that
more walls be detailed with Special Boundary Elements
(SBE) by adding a 1.5 factor in the denominator of ACI 318-
14 Equation 18.10.6.2. These more recent changes were a
result of the unsatisfactory performance of many walls in the
2010 Chile and 2011 New Zealand earthquakes, as well as
observations from recent large-scale laboratory tests
(Wallace, 2012; Wallace et al., 2012; Nagae et al., 2011;
Lowes et al., 2012).
Even with these more stringent requirements, the underlying
premise of the ACI 318-14 and ASCE 7-10 approach to
design and detailing Special Structural Walls is that walls
satisfying the provisions of ACI 318-14 §18.10.6.2 through
§18.10.6.4 possess adequate displacement capacity to exceed
the expected displacement demands from ASCE 7-10
analysis procedures when subjected to design-level ground
motions, without significant strength degradation. However,
recent research has shown that drift capacity of code-
compliant walls is impacted by parameters that are not
adequately addressed in ACI 318-14. For example, the
studies by Abdullah and Wallace (2018a) and Segura and
Wallace (2018b) showed that lateral drift capacity of walls
with SBEs is significantly influenced by parameters, such as
width of flexural compression zone b, wall length lw, neutral
axis depth c (i.e., compression demands), wall shear demand
1
 50 50 80 60 50 (a) 50 (b)

No. of Specimens
No. of Specimens
40 40 40 40
60
40 30 30
30 30
40
20 20 20 20
Vu, and configuration of boundary transverse reinforcement 20
wall provisions 20 of ACI 318 code do not adequately address
10 10
(overlapping hoops versus a single perimeter 10hoop with
10
safety concerns, nor do they ensure that walls have adequate
0 0
intermediate crossties, a detail0
allowed by ACI 0
318-14 0
drift capacity to meet the expected drift demands under DE
0

10
8

0. 9
1. 0

0.
3

0.
4
7

8
8

0
0. 9

0.
1

0
0

1-
2

4-
3

6-
5

3-
3

3.

5-
5

7-
0.

10
7

2.
4

0.
8

1.

7-

8-
0.

-3
5

-2

-4

2.

3.

9-
5-

5-
1.

5-
1-

4-

6-
3-

5-

7-

0-
7-

5-

0-
8-

0-

0.
§18.10.6.4). The findings of these studies indicated that, shaking with a reasonable
M/(Vl ) level of reliability (e.g.,
sv/db90%);

5-
9-

0.
20

2-

3-
10

30

1-

1.

2.

3.
2.
0.

1.
0.
sv/db Ash, prov./A P/(f A ) (%)
' Min. Ash, provide
sh,req: X-Dir. c g w

depending on these variables,60 drift capacity of60walls with 60 therefore, ACI 318-14
60 wall provisions should 60 be updated to
(e) 50
(f)

No. of Specimens
SBEs varies by a factor of about three, ranging between address this critical issue.
No. of Specimens

40
40
approximately 1.2% and 3.5%,40as shown in Figure 40 1. Figure 40 40 30

1 (a) presents variation of drift capacity of a dataset of 164 To address the above issue, a new reliability-based design 20
20 20
generally code-compliant walls versus a 20slenderness 20
methodology is 20proposed where a drift demand-to-capacity 10

parameter (λ b )
= lw c b 2 , whereas
0 Figure 1 (b)
0 presents 0
ratio (DDCR) 0check is performed to provide 0 a low
probability that roof drift demands exceed roof drift capacity
0

10

0. .5
0
5

0. 65
5
12 5

20 0

0. 3
15 0

0
10

-2
5

-1
0

10 0

<5

7.
5

-0
12

20

0.
0

0
15

5
30

5-
0
00
-1

0.
-2

0
<5

7.

( )

-2
2

-4
4

15
-1

-3
3

10
5-
5-

<1

1-

25
-1

7.
0-

4-
0-

5-
1-

3-
10

5-

0-

2-
15

>3
5-

15

30
10

20
7.

6
0.
results of the 121 slender walls M Vlwl /b≥ 2.0 in the dataset. 90
l /b hx/
b (mm) w at strength
c/b loss for a given
c/l hazard level (e.g., 10% probability
w
w

of lateral strength loss for the DE or MCE level shaking). In

These results have very important design implications. First,
not all code-compliant walls have the same, or nearly the general, walls with slender cross sections lw b > 15 , large ( )
same, drift capacity. Second, at Design Earthquake (DE) neutral axis depth relative to width of flexural compression
level shaking, ASCE 7-10 §12.12.1 limits allowable story
drift ratio to 0.02 for typical RC buildings in Risk Category I ( )
zone c b > 3 , shear stress demands approaching the ACI
& II that are taller than four stories and utilize structural walls 318-14 §18.10.4.4 limit of 10 f 'c , and roof drift demands
for a lateral force resisting system (LFRS). At Maximum
Considered Earthquake (MCE) level shaking, which is used approaching the maximum value allowed by ASCE 7-10 (i.e.,
to assess collapse prevention, this limit is typically taken as 0.75 × 0.02 = 0.015) tend to be screened out for redesign to
1.5 times the DE limit, or 0.03. If roof drift demand is prevent strength loss under DE level shaking and reduce the
approximated as three-quarters of peak story drift, which is a probability of collapse under MCE level shaking. This
reasonable approximation for buildings with walls, then the approach is rational and transparent and is similar to other
peak roof drift demand allowed by ASCE 7-10 is checks for moment and shear strength found in the ACI 318
approximately 0.0225 for MCE level shaking, which is about code. Finally, two design examples are presented to highlight
87% greater than the minimum wall drift capacity of 0.012 the deficiencies in the current code provisions and to
observed in Figure 1. These findings suggest that current illustrate the application of the proposed methodology.

5
vmax/÷ fc' psi £ 5 (£ 0.42 in MPa) vmax/÷ fc' psi £ 5 (£ 0.42 in MPa)
4 vmax/÷ fc' psi > 5 (> 0.42 in MPa) vmax/÷ fc' psi > 5 (> 0.42 in MPa)
Drift Capacity (%)

0
0 40 80 120 0 40 80 120
lb = lwc/b2 lb = lwc/b2
(a) Entire dataset (M/Vlw (b) M/Vlw .0 1.0)
Figure 1–Drift capacity variation of walls with special boundary elements (Abdullah and Wallace, 2018a).

5
)
Existing Design Approach
ACI 318-14 §18.10.6 includes two design approaches to
evaluate the need for SBEs at wall boundaries, a simplified
displacement-based design approach (§18.10.6.2) for slender
( )
walls hw lw ≥ 2.0 and a stress-based approach (§18.10.6.3).
Since the present study focuses on a drift demand-to-capacity
ratio (DDCR) approach for more slender walls with a single
critical section, the discussion that follows is limited to the
displacement-based design approach of §18.10.6.2, which
(
applies to slender walls hw lw ≥ 2.0 that are effectively
continuous from a single critical section to the top of the wall.
Detailed background information on the simplified
displacement-based design approach, which was first adopted
in ACI 318-99, is provided by Wallace and Orakcal (2002).
Per this approach, if the maximum value of depth of neutral
( )
axis cmax computed for the factored axial load Pu,max and
nominal moment strength (M ) ,
n
direction of the design roof displacement δ u , exceeds
critical depth of neutral axis ccritical ( )
(ACI 318-14 Equation 18.10.6.2), then SBEs are required at
the wall boundary regions over a distance greater than, or
equal to, the maximum of c 2 and c − 0.1lw . The 1.5
multiplier on δ u was added in ACI 318-14 to account for
dispersion in the computed drift demands under DE level
shaking and to produce detailing requirements more
consistent with the ASCE 7 code intent of a low probability
of collapse for MCE level shaking.
lw
ccritical =
(
600 1.5δ u hw ) δc
If a structural wall is determined to require a SBE based on
Eq. 1, then the SBE is required to satisfy the detailing
requirements of ACI 318-14 §18.10.6.4. If these detailing
requirements are satisfied, the underlying premise of the code
is that the wall drift capacity exceeds the expected wall drift
demands determined from analysis when subjected to DE-
level ground motions, without critical strength loss. However,
as presented earlier, this is not always the case. In particular,
walls with lw c b 2 > 45 and high shear stresses (e.g.,
approaching the ACI 318-14 §18.10.4.4 average wall shear
stress limit of 10 f 'c ( psi) ) would be expected to have a
( )
mean roof drift capacity δ c hw less than the maximum roof
drift demand allowed by ASCE 7-10 for MCE level shaking
(i.e., 0.03 × 0.75 = 0.0225), as can be seen in Figure 1. Walls
with these attributes are fairly common in modern wall
buildings (Brown et al., 2006). Therefore, the current design
provisions do not provide adequate safety margins. Next, a
new reliability-based design approach is proposed, which has
been proposed to ACI 318-19 code to address this issue.
Proposed Approach
A rational and transparent approach to address the
deficiencies highlighted in the previous section is to introduce
) a DDCR check for Special Structural Walls, which is
somewhat similar to demand-to-capacity checks in ACI 318
code for moment and shear strengths, or drift capacity of
slab-column connections (ACI 318-14 §18.14.5), to meet a
specified level of reliability. The basis for the new design
approach is expressed in Eq. 2:
( ) ⎛ δ ⎞ ⎛ 1.5δ u ⎞
consistent with the φd ⎜ c ⎟ ≥ ⎜ ⎟ (Eq. 2)
⎝ hw ⎠ ⎝ hw ⎠
( )
computed from Eq. 1 Where δ c hw is the mean wall lateral drift capacity
estimated from Eq. 3, a drift capacity model developed by
Abdullah and Wallace (2018a, 2018b) with a mean of 1.0 and
a coefficient of variation (COV) of 0.15, φ d is a
“displacement” reduction factor which can vary depending on
the level of reliability required, and δ u hw is the mean roof
drift demand estimated using ASCE 7 analysis approaches,
multiplied by a 1.5 factor to convert the DE drift demands to
MCE drift demands (ASCE 7-10 §11.4.4). This format also
is consistent with 1.5 multiplier used in the current ACI 318-
14 Equation 18.10.6.2 (or Eq. 1 herein).
(Eq. 1)
(
l c
% ) = 3.85 − w 2 −
vmax
≥ 1.5% (Eq. 3)
hw 50b 10 f 'c ( psi)
Based on results of nonlinear response history analyses (NL-
RHA) of 28 buildings with planar structural walls (Wallace
and Safdari, 2018), as well as results reported in the literature
(e.g., Kim 2016), Abdullah and Wallace (2018c) concluded
that the COV for roof drift demand under DE level shaking
generally ranges from 0.20 to 0.40. A COV of 0.30 was
adopted for the reliability analysis presented here, and the
sensitivity of the results to modest variations in the COVs
selected is considered later.
A simple reliability analysis of Eq. 2, assuming lognormal
distributions in roof drift demand and capacity with the
estimated COVs on drift demand and capacity and taking φ d
= 1.0, results in a probability of strength loss of
3
approximately 10% and 50% for DE and MCE level Abdullah and Wallace (2018c) reviewed results of wall tests
demands, respectively, as shown in Figure 2. If the COVs on with λb > 40 (i.e., lower drift capacities) in a dataset of 164
roof drift demand and capacity are increased to 0.40 and 0.2,
generally code-compliant walls and observed that lateral
respectively, the probability of strength loss under DE
strength loss in these walls was abrupt and typically much
demands increases modestly from about 10% to 17%,
greater than 20%, and that axial failure was observed to occur
indicating the strength loss probabilities are not overly
sensitive to the estimated COVs. These levels of probability soon after loss of lateral strength, even for walls with
of strength loss appear to be high, given the target collapse restrained web longitudinal bars (i.e., Segura and Wallace,
2018a; Shegay et al., 2016). Although tests of well-detailed,
probabilities of ASCE 7-16 §1.3.1.3 of 10% for Risk
isolated cantilever walls show that axial failure may follow
Category I and II buildings and 5% for Risk Category III
soon after substantial lateral strength loss under continued
buildings under MCE level demands. To reduce the
lateral loading, collapse of buildings with structural walls has
probability of strength loss to 10% for MCE level demands, a
rarely been reported following earthquakes or shake table
φ d of 0.65 is required. Selection of an appropriate φ d value
tests, even for walls with substantial damage (Wallace et al.,
requires a definition for collapse, since drift capacity at 20%
2008; Nagae et al., 2015). Given these observations, use of a
strength loss, which is used by Eq. 3, is not necessarily
low probability of strength loss (i.e., 10%) for DE level
associated with building collapse, which is more commonly
shaking is suggested here as a minimum criterion for collapse
associated with loss of axial load capacity. Use of a low
probability (10%) of strength loss for MCE level demands (i.e., φ d = 1.0). This approach will screen out walls with high
likelihood of strength loss at DE level shaking for redesign,
would be a conservative estimate of collapse, since axial
failure models in the literature for columns (Elwood and which will reduce the likelihood of severe damage at hazard
Moehle, 2005) and for walls (Wallace et al., 2008), as well as levels less than DE level shaking and reduce the potential for
collapse at MCE level shaking. Future adjustment to this
ASCE 41 backbone relations, generally indicate that drift
relation could be considered once additional information
ratios at axial failure exceed those at significant strength loss.
becomes available.

Figure 2–Probabilities of strength loss under DE and MCE hazard levels.

alternative approach would be to determine the required

If Eq. 2 is not satisfied for a given wall, then the designer
would be required to revise the design for that wall. The most
minimum width of flexural compression zone bmin by ( )
likely change would be to increase the width of the flexural rearranging Eq. 2 after substituting Eq. 3:
lw c
(b )
compression zone b (i.e., wall thickness, tw), which would 2
increase the drift capacity obtained with Eq. 3 by reducing ≥ (Eq. 4)
min
⎛ vmax 1.5δ u ⎞
the slenderness parameter λb = lw c b 2 and also likely 50 ⎜ 3.85 − − ⎟
reducing the shear and drift demands by increasing the shear ⎜⎝ 10 f c' (psi) hw ⎟⎠
resisting area and effective stiffness, respectively. This may
require some iteration; however, a more convenient

4
Where 1.5 δ u hw should not be taken less than the lower
limit on Eq. 3, which is 1.5%. An upper-bound width of the
( )
flexural compression zone bupper can be approximated using
Eq. 4, which is based on assuming the shear stress term
The discussion and results presented here are intended to
apply to walls with rectangular, flanged, and barbell cross
sectional shapes (Fig. 3(a) through (f)). For cases with a
large b, e.g., where the barbell or flange of the wall is in
compression (Fig. 3(b) through (h)), drift capacity is likely to

(v max
10 f 'c ) approaches 1.0 and roof drift demand term be relatively large (due to low λb ); however, for cases with a
barbell or flange in tension, and a thin wall web in
(1.5δ u hw ) approaches 0.0225 corresponding to the compression (Fig. 3(b) and (e) through (h)), relatively large
values of c b , and thus λb , and higher shear demands are
allowable roof drift demand under MCE level shaking,
resulting in the following: likely, and thus, lower drift capacities will result. Therefore,
such cases are more critical than cases where the barbell or
flange of the wall is in compression. For cases where b varies
bupper = 0.03clw (Eq. 5) over c, or where c varies over b, a representative (e.g.,
weighted average) value of b or c should be used, as shown in
Note that, if c = 0.20lw, then Eq. 5 requires bupper of 18.6 in. Fig. 3(c), (d), (e) and (h).

(470 mm) and 28 in. (710 mm) for walls with length of 20 ft
(6096 mm) and 30 ft (9144 mm), respectively.

Fig. 3–Definition of width b ( ) ()

and length c of flexural compression zone.

5
Alternative Approaches
1. Stringent Detailing Requirements
An alternative strategy that might be suggested to address the
deficiencies identified previously would be to require
sufficient detailing such that all walls have a roof drift
capacity that exceeds a “worst-case” for a story drift demand
of 0.03 for MCE level demands (a roof level demand of
approximately 0.0225). For example, Welt et al. (2018),
using results of rectangular prism tests, recommend
significantly stricter detailing requirements than what is
currently required by ACI 318-14, such as tighter spacing of
transverse and longitudinal reinforcement, restraint of every
longitudinal bar, and an increase of the amount of transverse
reinforcement. Additionally, Behrouzi et al. (2017)
recommend that the boundary element confinement pattern
and horizontal spacing and lateral restraint pattern of
boundary longitudinal reinforcement should follow the
provisions for high axial-stress columns, specifically
provisions of ACI 318-14 §18.7.5.2, which requires every
boundary longitudinal bars be restrained and a limit of 8 in on
the horizontal spacing of longitudinal bars. This approach
was used recently to update column detailing requirements in
ACI 318-14 §18.7.5.4 (Elwood et al., 2009); however, this
approach would be overly conservative for structural walls
where story drift demands are often considerably less than
0.03, e.g., for a building with many walls. Furthermore,
Abdullah and Wallace (2018a) studied the impact of detailing
variables on wall lateral drift capacity and concluded that
requiring more stringent detailing provisions do not
significantly improve drift capacity in many cases.
Parameters considered included: (1) Minimum ratio of
provided-to-required (per ACI 318-14 §18.10.6.4) area of
boundary transverse reinforcement, Ash, provided Ash,required , (2)
Ratio of vertical spacing of boundary transverse
reinforcement to minimum diameter of longitudinal boundary
reinforcement, s db , (3) Centerline distance between
laterally supported boundary longitudinal bars normalized by
maximum spacing, hx hmax , or by width of compression
zone, hx b , (4) degree of lateral restraint provided (i.e.,
support for all boundary longitudinal bars versus every other
bar). For these variables, they utilized a dataset of 78 wall
tests that fully satisfied the ACI 318-14 provisions,
particularly those related to quantities Ash, provided , s , s db ,
hx , and length of confined boundary element (l ) .
be The
results are presented in Fig. 4. It can be noted from this figure
that changes in these parameters within ranges that are
permissible or reasonable for SBEs do not influence (or
6
change significantly) the lateral drift capacity. This is in part
because it is difficult to get more drift capacity out of thin
walls because of lateral instability problems, as highlighted
by Segura and Wallace (2018a).
2. Limits on Wall Axial Load
Another approach that has been suggested is to place a limit
on axial load ratio (Shegay et al., 2018). This is because axial
load has traditionally been assumed to have a significant
impact on wall (and column) lateral drift (or plastic rotation)
capacity. For example, in UBC 1997 §1921.6.6.3 and ASCE
41-13 §10.7.1.1, if axial load on a wall exceeded 0.3 Ag f 'c ,
the lateral strength and stiffness of the wall could not be
considered. Additionally, ASCE 41-13 Tables 10-19 and 10-
20 use axial load ratio as a primary term for selecting
modeling parameters and acceptance criteria for both flexure-
and shear-controlled walls. Similar limits (0.3 or 0.35 Ag f 'c )
on wall axial load exist in other codes such as NZS 2017,
CSA 2004, and DS 60 MINVU (Chilean Structural Standard
2010). However, results presented by Abdullah and Wallace
( )
(2018a) revealed that axial load ratio P Ag f 'c , by itself,
has no clear correlation with wall drift capacity (correlation
coefficient, R = 0.08), as shown in Fig. 5. From this figure, it
is clear that there is no significant trend between axial load
ratio (ranging from 0.0 to 0.35) and wall drift capacity. It is
noted that the slenderness parameter λb described earlier
incorporates the impact of axial load through neutral axis
depth. Therefore, placing a limit on axial load is not
recommended.
3. Change to ASCE 7 Provisions
Lastly, one might suggest that the wall provisions of ASCE 7
could be changed to address the issues of inadequate
deformation capacity of walls, particularly related to reducing
the seismic response coefficient ( )
R , deflection
( )
amplification factor C d , and/or the allowable story drift
(Δ a )
hx . However, this approach would also be too
conservative because it would punish all walls, even walls
with small λb and shear stresses which tend to possess quite
large drift capacities, as was seen in Figure 1.
For the reasons mentioned here, the above three approaches
were not considered to be effective and, thus, are not
recommended. Next, two design examples are presented to
illustrate the proposed DDCR approach.
 0 40 80 120 0 40 80 120
lwc/b2 lwc/b2

5
1.0 Ash,provided/Ash,required 1.2 1.5 s/db < 3.5
4 1.2 < Ash,provided/Ash,required 3.5 < s/db 6
Drift Capacity (%)

1
a) BE transverse reinf.: X-Dir b) Bar slenderness ratio
0
5
All bars supported 0.3 < hx/hx,max
4 Not all bars supported 0.7 < hx/hx,max
Drift Capacity (%)

1
d) Spacing of laterallysupported BE
c) Lateral support of BE longi. bars longi. bars
0
0 40 80 120 0 40 80 120
lb = lwc/b2 lb = lwc/b2
Figure 4–Impact of boundary element details on drift capacity of walls with SBEs (Abdullah and Wallace, 2018a).

Figure 5–Impact of axial load ratio on drift capacity of walls with SBEs (Abdullah and Wallace, 2018a).

7
Design Examples concrete compressive strength (f' )
c of 5 ksi and Grade 60

Description of the Buildings reinforcement with yield strength (f )

y
of 60 ksi are
In the following, two residential buildings (6-story and 10-
specified, consistent with requirements of ACI 318-14
story) located in Los Angeles, California are used to illustrate
§18.2.5 for concrete and §18.2.5 for reinforcement in
the application of the proposed design methodology, as well
structural walls. A total uniformly distributed floor dead load
as to highlight the significant deficiency in the current design
(in addition to self-weight of walls) of 150 psf and floor live
provisions of ACI 318. The building footprint is 150×75 ft.
load of 40 psf per ASCE 7-10 §4.3.1 are used as the loading
(Fig. 6), and the typical story height is 12 ft. A summary of
criteria.
seismic design parameters is provided in Table 1. Design

6@25ft (7.6m) = 150ft (45.6m)

D
3@25ft (7.6m) = 75ft (22.8m)

Tributary Area

Wall #2
Wall #1

N
B
24"x24"
Gravity column (typ.)
A

1 2 3 4 5 6 7
Figure 6–Typical plan view of the buildings.
the building.
(a) Plan view of
considered. A wall effective stiffness, Ec I eff , of 0.5Ec I g
Table 1–ASCE 7-10 seismic parameters
was assumed for the lateral analysis, consistent with ACI
Parameter Value 318-14 §6.6.3.1.2. The contribution of the gravity columns to
Building Location 34.058°N, 118.445°W the lateral strength and stiffness of the system was ignored.
The lateral analysis included the impact of accidental
Risk Category II
Importance Factor 1.0
( )
torsional moment M ta required by ASCE 7-10 §12.8.4.2.
Inclusion of accidental torsion generally resulted in an
Site Class D
increase of both roof drift and base shear demands by about
SS; SDS (g) 2.253; 1.502 15%. The ASCE 7-10 strength level load combinations (LC)
S1; SD1 (g) 0.829; 0.829 defined in §2.3.2 and §12.14.3.1 were used to compute the
ultimate force demands. Additionally, a redundancy factor
Seismic design category (SDC) E
R = Cd 5
( )
ρ of 1.3 was applied to the load combinations that include
seismic loads in accordance with ASCE 7-10 §12.3.4.2,
Redundancy factor ρ 1.3 resulting in a 30%-increase in base shear and moment
demands, and a ρ of 1.0 was used for drift calculations in
accordance with ASCE 7-10 § 12.3.4.1. It its noted that
Lateral Load Analysis
ASCE 7-10 §12.9 Modal Response Spectrum Analysis (RSA) ( )
1.2 + 0.2S DS D + 0.5L + ρQE load combination, with
was utilized to determine design lateral forces on the walls negative accidental eccentricity (i.e., moving center of mass
under DE level shaking. For the purposes of this study, only (CM) closer to the wall), produces in the largest force and
analysis and design of the LFRS in the north-south direction, drift demands.
which consists of planar RC Special Structural Walls, was

8
Detailed information for the LFRS and the analysis results
(maximum story and roof drifts, base moment, and base shear
demands) are summarized in Table 2 under columns A6 and
A10 for the 6-story and 10-story buildings, respectively. The
walls were proportioned such that the maximum story drift
demands (Δ story )
hx , computed at CM in accordance with
ASCE 7-10 §12.8.6, were smaller than the allowable story
drift (i.e., Δ a hx = 0 .02) given in ASCE 7-10 §12.12.1 for
( )
DE level shaking. The DE roof drift demands δ u hx , given
in Table 2, were taken at the top of the wall (not at CM),
consistent with wall design displacements used in ACI 318-
14 Equation 18.10.6.2. A factor of 1.5 was used to convert
the DE roof drift demands to MCE demands, as noted
previously. Base moment, shear, and axial demands given in
Table 2 are for a critical section at the base of the walls.
Walls Design
Based on the demands from the preceding section, two
identical planar structural walls are proposed as the LFRS in
the north-south direction for each building. The walls are 24
ft. long and 12 in. thick for building A6 and 26 ft. long and
18 in. thick for building A10, resulting in wall cross-section
( )
aspect ratio lw b of 24 and 17.33, respectively. Because
the buildings are assigned to SDC E in accordance with
ASCE 7-10 §11.6, the walls are required to be designed and
detailed to satisfy the provisions of ACI 318-14 §18.10 for
Special Structural Walls. Wall design details are shown in
Table 2 under columns 6A and 10A for the 6-story and 10-
story buildings, respectively. Since cmax > ccritical , the
compression zones of the walls must be reinforced with SBE
details that satisfy the requirements of ACI 318-14 §18.10.6.4
( )
over a distance of lbe,required = max c 2;c − 0.1lw .
Reliability Analysis
Wall roof drift capacities given in Table 2 were computed
from Eq. 3, which are defined at 20% strength degradation.
The lower bound (minimum) drift capacity from Eq. 3
governs for walls in both 6A and 10A buildings. To
determine the probabilities of strength loss, simple reliability
analyses of Eq. 2 were performed assuming lognormal
distributions in drift demand and capacity and considering
COVs of 0.30 and 0.15 for roof drift demand and capacity,
respectively. The probabilities of strength loss under DE and
MCE level shaking are given in Table 2. The resulting
values of 45% and 86% for DE and MCE level shaking,
respectively, for building 6A, and 65% and 94% for DE and
MCE level shaking, respectively, for building 10A are
unacceptably high given the current target reliabilities of
ASCE 7-16 §1.3.1.3. It is important to note that the walls in
both buildings (6A and 10A) satisfy the provisions of ASCE
7-10 and ACI 318-14 for Special Structural Walls (i.e., code
compliant walls). These results highlight that the current code
provisions do not provide adequate safety margins, and that
these wall designs should not be accepted.
Revised Design
To reduce the probability of strength loss to an acceptable
level (e.g., 10% or lower for DE level shaking), either Eq. 4
( ) ( )
bmin or Eq. 5 bupper can be employed. For the given c lw
(
demands, the upper bound compression zone width bupper is )
27.8 in. for the 6-story building and 30.5 in. for the 10-story
building. An alternative approach is used here, where bmin is
determined using Eq. 4 assuming a change in wall thickness
results in proportional reductions in δ u hw , v max f 'c , and
c lw . For building 6A, revised demand values for an
estimated 15% reduction are: δ u hw ≈ 1.44×0.85 = 1.22%,
v max f 'c ≈ 7.21×0.85 = 6.13, and c lw ≈ 0.31×0.85 =
0.26.
Substituting these values in Eq. 4 results in bmin = 0.061 lw ≈
18 in. for the 6-story building and, similarly, bmin = 23 in. for
the 10-story building. Therefore, wall thickness values were
increased to 18 in. for the 6-story building and to 24 in. for
10-story building. Using the new wall thickness values, the
analyses were rerun to determine the new force and drift
demands, as well as to determine whether Eq. 2 is satisfied
(i.e., probability of strength loss is 10% or lower for DE level
shaking). The revised design details are given in Table 2
under columns 6B and 10B for the 6-story and 10-story
building, respectively. As can be seen from Table 2, increase
in the wall thickness for building 6A resulted in: (1) reduction
of δ u hw by about 17%, (2) reduction of v max f 'c by
about 16%, and (3) significant increase in δ c hw , because a
portion of the drift capacity is proportional to b 2 . The new
probabilities of strength loss for DE and MCE level shaking
have reduced to below 10% and 50%, respectively, as shown
in Table 2.
9
 Table 2–Design details of the walls in each building
6-story 10-story
Building
6A 6B 10A 10B
hw × lw × t w (ft) 72×24×1.0 72×24×1.5 120×26×1.5 120×26×2.0

(
hw lw ; M base Vbase lw ) 3.0; 2.02 4.62; 2.39
lbe × b (in.) 61×12 52×18 71×18 53×24
lbe,required (in.) 60.2 41.2 69.8 50
Boundary longitudinal
45 No.11 45 No.11 42 No.14 44 No.14
reinforcement
Boundary transverse
No.4@3.75in. No.4@4.5in. No.5@4in. No.5@4.5in.
reinforcement
Ash, prov . Ash,req. ;S prov . Sreq. 1.05; 0.9 1.02; 0.8 1.00; 0.7 1.05; 0.8
Web vertical and horizontal 2layers
2 layers No.6@8in. 2 layers No.6@9in. 2 layers No.6@10in.
reinforcement No.6@8.5in.
Min φ M n (kips-ft.) 88,139 95,570 133,039 152,076

V@ Mn (A cv )
f 'c (psi) 8.3 6.0 6.0 5.1

Vn,ACI (A cv
f ' (psi) )
c
9.8 6.6 7.2 5.7

Ta ;Tu ;T1 (sec) 0.49; 0.69; 0.94 0.49; 0.69; 0.78 0.73; 1.02; 1.79 0.73; 1.02; 0.1.58
Pu1 (kips); Pu1 (A f ' )
g c 2,420; 0.14 2,649; 0.10 4,606; 0.16 5,022; 0.13

Pu2 (kips); Pu2 (A f ' )

g c 901; 0.05 991; 0.038 1,724; 0.06 1,888; 0.05

M base (kips-ft.) 85,356 92,050 132,213 146,073

Vbase (kips) 1,762 1,835 2,127 2,222

Vbase (A cv
f 'c (psi) ) 7.21 6.0 5.36 4.2

cmax ; ccritical (in.) 89; 22.2 70; 26.7 101; 20.3 81; 23

Max c lw 0.31 0.24 0.32 0.26

c b ; lw b 7.5; 24 3.89; 16 5.6; 17.33 3.38; 13

λb = l w c b 2
180 62 97 44

Max Δ story hx (%) at DE 1.92 1.56 1.93 1.70

ASCE 7-10 Δ a hx (%) 2.00 2.00 2.00 2.00
Roof δ u hw (%) at DE 1.44 1.20 1.71 1.51
Roof δ u hw (%) at MCE 1.44×1.5 = 2.16 1.20×1.5 = 1.80 1.71×1.5 = 2.56 1.51×1.5 = 2.27
δ c hw (%) (Eq. 3) 1.5 2.01 1.5 2.55
Probability of strength loss (%) at
45 (86) 6 (36) 65 (94) 6 (36)
DE (MCE)

10
Conclusions and Recommendations
Based on the findings of this study, the following conclusions
and recommendations with regards to design of structural
walls with SBEs resulted:
1. The underlying premise of the ASCE 7-10 and ACI 318-
14 provisions is that special structural walls satisfying the
provisions of ACI 318-14 §18.10.6.2 through §18.10.6.4
possess adequate drift capacity to exceed the expected
drift demand determined from ASCE 7-10 analysis
procedures. However, results presented in this study show
that these code provisions are insufficient to provide
ductile, flexure-dominated response under cyclic loading,
and, thus, do not provide adequate safety margins for wall
buildings.
2. To address the above deficiencies, a new reliability-based
design methodology is proposed where a drift demand-to-
capacity ratio (DDCR) check is performed to provide a
low probability (i.e., 10% or lower) that roof drift
demands exceed roof drift capacity at strength loss for the
DE level shaking, which is a rational and transparent
approach and can easily be implemented in ACI 318.
(
3. In general, walls with slender cross sections lw b > 15 , )
large neutral axis depth relative to width of flexural
( )
compression zone c b > 3 , shear stresses approaching
( )
the ACI 318 §18.10.4.4 limit 10 f 'c (psi) , and roof
drift demands approaching the maximum story drift
allowed by ASCE 7-10 are screened out for redesign.
4. Preventing strength loss under DE level shaking is
assumed to reduce the probability of collapse under MCE
level shaking; however, for improved performance, a
lower (or specified) probability of strength loss for MCE
level shaking could be used. That is, a lower displacement
reduction factor (φ d < 1.0) could be used in Eq. 2.
5. Alternative approaches to address the noted concerns,
such as requiring more strict detailing provisions than
what are currently required by ACI 318-14, placing a limit
on axial load ratio, and/or modifying wall provisions of
ASCE 7, are not recommended.
6. Example applications are presented to highlight current
code deficiencies and application of the proposed
approach. To assist in cases where redesign is required,
expressions for minimum and upper-bound width of
flexural compression zone are provided.
References
ACI Committee 318, 2014, “Building Code Requirements for
Structural Concrete (ACI 318-14) and Commentary,”
American Concrete Institute, Farmington Hills, MI, 519 pp.
Abdullah S. A., Wallace J. W., 2018a, “Drift capacity of RC
structural walls with special boundary elements,” ACI
Structural Journal, accepted for publication, 33pp.
Abdullah, S. A., and Wallace, J. W., 2018b, “Drift capacity
prediction of RC structural walls with special boundary
elements,” Proceedings, 11th National Conference in
Earthquake Engineering, Earthquake Engineering Research
Institute, Los Angeles, CA.
Abdullah S. A., Wallace J. W., 2018c, “A reliability-based
design methodology for RC structural walls with special
boundary elements,” ACI Structural Journal, submitted for
review and possible publication, May 4, 2018, 33pp.
ASCE/SEI Standards, 2013, “Seismic Evaluation and
Retrofit of Existing Buildings (ASCE/SEI 41-13),” American
Society of Civil Engineers, Reston, VA, 518 pp.
ASCE/SEI Standards, 2010, “Minimum Design Loads for
Buildings and Other Structures (ASCE/SEI 7-10),” American
Society for Civil Engineers, Reston, VA, 518 pp.
ASCE/SEI Standards, 2016, “Minimum Design Loads for
Buildings and Other Structures (ASCE/SEI 7-16),” American
Society for Civil Engineers, Reston, VA, 690 pp.
Behrouzi, A., Welt, T., Lehman, D., Lowes, L., LaFave, J.,
and Kuchma, D., 2017, “Experimental and numerical
investigation of flexural concrete wall design details,”
Proceedings, ASCE/SEI Structures Congress, Denver,
Colorado.
Brown, P., Ji, J., Sterns, A., Lehman, D. E., Lowes, Kuchma,
D., and Zhang, J., 2006 “Investigation of the seismic
behavior and analysis of reinforced concrete walls,”
Proceedings, 8th National Conference on Earthquake
Engineering, San Francisco, CA.
Chilean Structural Standard, 2010, “Reinforced concrete
design code, replacing D.S N 118 (DS 60 MINVU),” Chilean
Ministry of Housing and Urbanism, Diario Official, Santiago,
Chile. (In Spanish)
Canadian Standards Association, 2004, “Design of concrete
structures (CSA A23.3-04), Canadian Standards Association,
Mississauga, ON, Canada.
Elwood, K. J., Maffei, J. M., Riederer, K. A., and Telleen,
K., 2009, “Improving Column Confinement–Part 2: Proposed
New Provisions for the ACI 318 Building Code,” Concrete
International, Vol. 31, No. 12, pp. 41–48.
11
Elwood, K. J. and Moehle, J. P., 2005, “Drift capacity of
reinforced concrete columns with light transverse
reinforcement,” Earthquake Spectra, Vol. 21, No. 1, pp. 71-
89.
Kim, S., 2016, “Reliability of structural wall shear design for
tall reinforced concrete core wall buildings,” Ph.D.
Dissertation, University of California, Los Angeles, CA, 246
pp.
Wallace, J. W., and Orakcal, K., 2002, “ACI 318-99
provisions for seismic design of structural walls,” ACI
Structural Journal, Vol. 99, No. 4, pp. 499–508.
Wallace, J. W, Massone, L. M., Bonelli, P., Dragovich, J.,
Lagos, R., Luders, C., and Moehle, J., 2012, “Damage and
implications for seismic design of RC structural wall
buildings,” Earthquake Spectra, Vol. 28, No. S1, pp. 281–
289.

Lowes, L. N., Lehman D. E., Birely A. C., Kuchma D. A., Wallace, J. W., Elwood, K. J., and Massone, L. M., 2008,
Marley K. P., and Hart C. R., 2012, “Earthquake response of “Investigation of the axial load capacity for lightly reinforced
slender concrete planar concrete walls with modern wall piers,” Journal of Structural Engineering, Vol. 134, No.
detailing” Engineering Structures, Vol. 34, pp. 455–465. 9, pp. 1548-1557.

Nagae, T., Tahara, K., Taiso, M., Shiohara, H., Kabeyasawa,
T., Kono, S., Nishiyama, M., Wallace, J. W., Ghannoum, W.
M., Moehle, J. P., Sause, R., Keller, W., and Tuna, Z., 2011,
“Design and Instrumentation of the 2010 E-Defense Four-
Story Reinforced Concrete and Post-Tensioned Concrete
Buildings,” PEER Report 2011/104, Pacific Earthquake
Engineering Research Center (PEER), Berkeley, CA, 234 pp.
Standards New Zealand, 2017, “Concrete structures standard.
Part 1: The design of concrete structures (Amendment No. 3)
(NZS 3101:2006)”, Standards New Zealand, Wellington,
New Zealand.
Segura, C. L., and Wallace, W. J., 2018a, “Seismic
performance limitations and detailing of slender RC walls,”
ACI Structural Journal, accepted for publication (minor
revisions), Oct. 24, 2017, 29 pp.
Wallace, J. W., and Safdari, A., 2018, “Design of slender
reinforced concrete walls,” Proceedings, 11th National
Conference in Earthquake Engineering, Earthquake
Engineering Research Institute, Los Angeles, CA. (final
paper accepted for publication, Paper ID 1861)
Welt, S., Lehman, D., and LaFave, J., 2018, “Boundary
element detailing in special concrete structural walls,” ACI
Structural Journal, Vol. 115, No. 3, pp. 635-647.
ICBO, 1997, Uniform Building Code, Structural Engineering
Provisions, Vol. 2, 1997 edition, pp. 2-161 to 2-163,
International Conference of Building Officials, Whittier,
California.

Segura, C. L., and Wallace, W. J., 2018b, “Impact of

geometry and detailing on drift capacity of slender walls,”
ACI Structural Journal, accepted for publication (minor
revisions), Dec. 1, 2017, 32 pp.

Shegay, A. S., Motter, C. M., Elwood, K. J., Henry, R. S.,

Lehman, D. E., and Lowes, L. N., 2016, “Impact of axial
load on the seismic response of rectangular walls,” Journal of
Structural Engineering, Vol. 144, No. 8, pp. 04018124-1
04018124-14.

Uniform Building Code, 1997, “International Council of

Building Code Officials (UBC-97),” Whittier, CA.

Wallace, J. W., 2012, “Behavior, design, and modeling of

structural walls and coupling beams– lessons from recent
laboratory tests and earthquakes,” International Journal of
Concrete Structures and Materials, Vol. 6, No. 1, pp. 3–18.

12

View publication stats