ACI 318-19 CH 011f – Wall Boundary Detailing

ACI 318 Change Submittal

Submittal #: CH 011f

Subject: Special Structural Walls – Wall Boundary Detailing

Code (or Commentary) Provisions: 18.10.6.4

Reason for Change: To address detailing deficiencies for slender (hw/lw ≥ 2) special
structural walls to require boundary vertical reinforcement be supported by overlapping
hoops and crossties with seismic hooks at each end and web vertical reinforcement be
supported by crossties with seismic hooks at each end over the height of the special boundary
element.

Ballot History:

Subcommittee (or Task Group or Main Committee) Ballot Results

Ballot No: CH-011 Canvas or meeting date:

(1) April 7, 2016

(2) March 13, 2017
(3) July 24, 2017
(4) October 10, 2017
Total number of eligible voters = 19

Code or Commentary Aff. Aff. w/ comment Neg. Abstain Not

provisions (Y) (Y/C) (N) (A) Returned

18.10.6.4a 8 2 7 0 2

18.10.6.4b 7 2 5 0 5

18.10.6.4c: July 24, 2017 9 2 7 0 0

18.10.6.4d: Oct. 10, 2017 9 1 6 0 0

18.10.6.4e: Feb. 12, 2018 22 5 10 3 0

LB18-01
ACI 318-19 CH 011f – Wall Boundary Detailing

Summary of Reasons for Negative Votes - CH-11e: No votes were, for the most part,
related to relatively minor (but substantive) changes to clarify and clean-up language, edit
notation for hx. Other significant change to address the no votes by Klein, Frosch, and
Seguirant was to separate the commentary figures for web anchorage and overlapping hoops.

Negative voters: Seguirant, Klorman, Parra, Klein, Kreger, Kelly, Wight, Sanders,
Rodriguez, Frosch (10).

Based on the changes, all negatives were withdrawn.

Background of Proposed Change:

Recent laboratory tests of special structural walls (Segura and Wallace, 2017a; 2017b) and
prisms (or rectangular columns) representing wall boundaries (Welt et al 2015; Arteta et al,
2014) have revealed that boundary elements with single perimeter hoops and supplemental
90-135 or 135-135 crossties do not always achieve the plastic deformation capacity implied
by ASCE 41 and are susceptible to rapid lateral strength loss and axial failure immediately
after lateral strength loss. Results for two tests are summarized in the following paragraphs
to highlight these issues. Additional information is provided in the referenced papers.

Test Specimens:
The experimental testing matrix for WP2 and WP4 is shown in Table 1. The primary variables were
the arrangement of longitudinal reinforcement (WP2) and compression depth (WP4). Boundary
transverse reinforcement was selected to satisfy ACI 318-14 minimum requirements for Special
Structural Walls (ACI 318 designation).

Table 1 – Specimen testing matrix

Ash,y / Ash,x / Ash,re q /sbc Ash /

s/db c/Lw
Test Specimen sbc,y sbc,x Eq. 2a Eq. 2b Ash,re q
--- --- [%] [%] [%] [%] ---
3.20
WP2 0.20 1.28 1.78 1.23 0.75 1.04
4.00
WP4 (web) 3.20 0.30 1.21 1.78 1.14 0.75 1.07

Cross-sectional geometries and reinforcement are shown in Fig. 1. Both test specimens were 2134mm
in height, 2286mm in length, and 152mm thick. Boundary longitudinal reinforcement for WP2
ACI 318-19 CH 011f – Wall Boundary Detailing

consisted of eight 15.9mm diameter bars at the west boundary (ρlb=2.92%) and fourteen 12.7mm
diameters bars at the east boundary (ρlb=3.27%). Web longitudinal and transverse reinforcement
consisted of 9.5mm diameter bars spaced at 203mm on-center (ρlw =ρlv=0.46%). Web longitudinal
bars were placed outside of web transverse bars (Fig. 1a) as allowed by ACI 318-14. Compression
depth, determined using an extreme fiber compression strain of 0.003 (consistent with ACI 318-14
requirement) for an axial load of 0.10Awf’c, was 0.21lw and 0.20lw for loading causing compression at
the west boundary and east boundary, respectively. For typical design roof drift demands (e.g.,
δu/hw=0.01-0.02), SBEs are required at both wall boundaries. SBE detailing at each boundary
consisted of an 8.1mm diameter hoop and two cross-ties spaced at 51mm on-center (s/db=3.2 and 4.0
at west and east boundaries, respectively). Boundary transverse reinforcement ratio (Ash/sbc) was
1.28%.

Specimen WP4 was constructed with an enlarged boundary region (flange) at the east end of the wall
(Fig. 1b). The flange accommodated more longitudinal reinforcement at the east boundary, making it
possible to impose a longer compression depth at the web boundary as would be expected for a wall
with higher axial load demand (e.g., 0.20Awf’c) or a wall with an asymmetric cross-section (e.g. T-
shape wall). Boundary longitudinal reinforcement for WP4 consisted of ten 15.9mm diameter bars at
the west (web) boundary (ρlb=2.84%) and fourteen 19.1mm diameters bars at the east (flange)
boundary. Web longitudinal and transverse reinforcement was identical to that of specimen WP2.
Compression depth, determined using an extreme fiber compression strain of 0.003 for an axial load
of 0.10Awf’c, was 0.30lw and 0.06lw for loading causing compression at the web and flange boundaries,
respectively. For typical roof drift demands, SBE detailing is required at the web boundary only. SBE
detailing at the web boundary consisted of an 8.1mm diameter hoop and three cross-ties spaced at
51mm on-center (s/db=3.2, Ash/sbc=1.21%). OBE detailing at the flange boundary consisted of two
overlapping 6.4mm diameter hoops spaced at 114mm on-center (s/db=6.0).
ACI 318-19 CH 011f – Wall Boundary Detailing

Ø9.5 @ 203

Hoop & cross-ties Hoop & cross-ties

Ø8.1 @ 51 Ø8.1 @ 51

8 Ø15.9 16 Ø9.5 14 Ø12.7

152 152

356 76 7@203 76 356

2286

a)

Ø9.5 @ 203

Hoop & cross-ties Hoops

Ø8.1 @ 51 Ø6.4 @ 114

10 Ø15.9 16 Ø9.5 14 Ø19.1

152 559

457 7@203 356

25 25
b) 2286

Fig. 1 – Specimen cross-sections: a) WP2; and b) WP4

Test Results
Normalized base overturning moment vs. hinge rotation results are shown in Fig. 2. Hinge rotation
was measured by two control sensors, positioned at opposite ends of the wall, over an assumed hinge
length of one-half the length of the wall. For each plot, the ASCE 41 modeling backbone is included.

Fig. 2 – Experimentally Measured Base Overturning Moment vs. Hinge Rotation

For specimen WP2, crushing of cover concrete at both ends of the wall was observed during the first
1.5% hinge rotation loading cycle. Minor crushing at the west boundary was observed during the next
ACI 318-19 CH 011f – Wall Boundary Detailing

loading cycle, resulting in a reduction in strength of approximately 30%. Lateral instability at the west
boundary was observed while attempting a third loading cycle to +1.5% hinge rotation. The
instability occurred at a small positive load, and little residual strength remained for loading in the
positive direction. Monotonic loading was applied in the negative direction until failure. At
approximately -1.7% hinge rotation, abrupt crushing at the east boundary occurred (Fig. 3a). Crushing
of concrete extended approximately one-third the length of the wall from the east boundary, and
damage was concentrated within the bottom 350mm above the base of the wall. Buckling of all
boundary longitudinal reinforcement and the four web longitudinal bars closest to the east boundary
was observed (Fig. 3b). Two fractured boundary hoops were removed from the specimen and opening
of several 90-degree cross-tie hooks was observed in the damaged region (Fig. 3c). Very little
residual lateral strength remained following the failure. The specimen was unable to achieve 2%
plastic rotation capacity at a residual strength of 75% of the nominal flexural strength, as assumed by
ASCE 41.

Fig. 3 – WP2: a) Damage at east edge; b) Buckled reinforcement; and c) Transverse reinforcement

For specimen WP4, crushing occurred at the web boundary at approximately +1.3% hinge rotation
with no prior signs of damage or strength loss. Damage concentrated within the bottom 350mm of the
ACI 318-19 CH 011f – Wall Boundary Detailing

wall, and extended horizontally approximately two-thirds the length of the wall. Out-of-plane
displacement of the compression zone occurred near the base of the wall, in the damaged region. A
maximum out-of-plane displacement of 60mm was measured approximately 300mm above the base
of the wall. Removal of lose concrete revealed buckling of several boundary longitudinal bars and a
few web longitudinal bars, fracture of boundary hoops, and failure of 90 degree cross-tie hooks. Like
specimen WP2, WP4 was unable to achieve the deformation capacity and residual strength assumed
by ASCE 41, and loss of axial load carrying capacity was lost immediately after loss of lateral
strength.

The detailed assessment of the test results indicates two primary issues. First, for a single
perimeter hoop, the long side parallel to the wall length lw is too flexible and does not
adequately restrain lateral expansion of the core concrete or the longitudinal reinforcement,
especially for longitudinal reinforcement not supported by a crosstie, and second, the 90-
degree legs of the crossties do not adequately restrain buckling of longitudinal reinforcement
once concrete spalling occurs. Once longitudinal reinforcement buckling initiates,
compressive strains concentrate over a wall height of approximately 2.5 times the wall
thickness, and significant strength loss is observed. Additional tests on SBEs subjected to
pure compression by Arteta and Moehle (2017) reveal similar results, and also indicate that
use of 135-135 crossties only marginally improves deformation capacity. In both test
programs, lateral instability of the damaged region was observed soon after compressive
strains concentrated.

Based on these observations, Segura and Wallace (2017a) conducted tests where the primary
test variables were the ratio of wall thickness to neutral axis depth, the use of one-piece (or
continuous) hoops/crossties (referred to here as Continuous Transverse Reinforcement, or
CTR), and the use of 135-135 crossties for web vertical reinforcement (Fig. 4) It is noted that
the CTR used in the test program was heat-treated, smooth reinforcement (Fig. 4b), given
that deformed bar was not available in small enough bar diameters for the one-half scale
tests. Load versus deformation responses for two tests are compared in Fig. 5 and indicate
that use of CTR lead to improved deformation capacity, and that the improved boundary and
web detailing produced less rapid strength loss and improved residual strength (Fig. 5).
ACI 318-19 CH 011f – Wall Boundary Detailing

Additional tests also revealed that increasing the b/c ratio lead to improved deformation
capacity.

15cm

(a) WP1 Cross Section

(b) WP5 Cross Section

Figure 4. Test wall cross sections

Figure 5. Moment versus hinge rotation relations for test walls WP1 and WP5.

Based on the test observations, a database was developed to compare the drift capacity of
walls with SBEs with and without overlapping hoops (Fig. 6). The data were filtered to
include only walls satisfying the following criteria:

(a) Test specimens subjected to quasi-static, cyclic loading,

(b) Measured concrete compressive strength ≥ 3 ksi (20 MPa),
(c) Width of compression zone, b ≥ 3.5” (90 mm) (most > 4 in.),
(d) Specimens with rectangular compression zone (except for 6 specimens),
(e) At least two curtains of web vertical and horizontal reinforcement,
(f) Specimens did not exhibit shear failure (i.e. diagonal tension, diagonal compression,
or sliding at the base) or splice failure,
ACI 318-19 CH 011f – Wall Boundary Detailing

(g) Ratio of ultimate to yield stress of boundary longitudinal reinforcement Fu Fy ≥ 1.2 ,

(h) Ratio of provided-to-required boundary transverse reinforcement
Ash , provided Ash ,required > 0.7 ,
(i) Ratio of vertical spacing of hoops/crossties of boundary transverse reinforcement to
boundary longitudinal bar diameter sv d b ≤ 7.0 (expect for four specimens).

Walls less than 3.5 inches thick were not included because use of two layers of web
reinforcement and realistic cover over SBEs is not practical. At least two curtains of web
reinforcement was specified to be consistent with current 18.10.2.2 requirements. The ratio
of ultimate-to-yield stress for boundary longitudinal reinforcement is slightly less restrictive
than requirements of 20.2.2.5, which requires a ratio of 1.25. The last two criteria, which are
slightly less restrictive than current 318 requirements, were used to include more data in the
assessment. A total of 164 tests were found to satisfy these criteria.

Data were split into two bins, as data analysis indicated that use of overlapping hoops

( )
improved drift capacity for moderate wall shear stress vu ≤ 5.5 f c' , provided the ratio of

neutral axis depth to wall thickness exceeded approximately 2.5, but not for higher wall shear

( )
stress vu > 5.5 f c' , and drift capacity improvement with overlapping hoops is significant

for a ratio of c/b > 4.0.

(a) Shear stress ≤ 5.5 f c' (b) Shear stress > 5.5 f c'

Figure 6. Drift capacity of walls with SBEs with and without overlapping hoops
ACI 318-19 CH 011f – Wall Boundary Detailing

The ACI 318-14 required depth of confinement, denoted c’ here, is the larger of c/2 and c-
0.1lw. If a neutral axis depth of 0.20lw is assumed for the database (which is about median
for the tests), then the required length of confinement for the tests summarized in Fig. 6 is: c’
= larger of (c/2; c-0.1lw = c/2) = c/2. It is noted that the average length of confinement for
the tests was about 20% greater than required, or c’ = 0.6c. Therefore, for the range of c/b
values where use of overlapping hoops improves deformation capacity (c/b > 2.5), the ratio
of the confined length to wall thickness is c’/b = 0.6(2.5) = 1.5, suggesting that the ratio of
the sides of a single perimeter hoop (denoted l1/l2 here) should not exceed about 1.5. For the
tests described early by Segura, where l1/l2 varied between 2.5 to 3.0, use of a single
perimeter hoop with crossties was ineffective. Given these observations, use of a ratio of l1/l2
> 2.0 is not recommended, and use a smaller ratio, e.g., 1.5, might be prudent. Finally, it is
noted that use of overlapping hoops was required by UBC-97 Section 19.21.6.6.6, for the
conditions noted in sub-sections 2.3 and 2.4.
2.3 The ratio of the length to the width of the hoops shall not exceed 3. All adjacent
hoops shall be overlapping.
2.4 Cross ties or legs of overlapping hoops shall not be spaced further apart than 12
inches (305 mm) along the wall.
Based on all of these factors, a ratio of 2.0 is recommended in this proposal. Two examples
of SBEs, one without (c = 2.5b) and one with (c = 5b) overlapping hoops, are shown in Fig.
7.
Perimeter hoop Supplemental crossties Vertical web Through web
reinforcement Crosstie

b bc2

bc1 (bc1 / bc2 ≤ 2) Horizontal web

reinforcement, Av

lbe

(a) SBE with single perimeter hoop and supplemental crossties

ACI 318-19 CH 011f – Wall Boundary Detailing

Hoop Overlap Horizontal web Through web

Supplemental crosstie at least min of reinforcement, Av
(6 in. and 2b/3 Crosstie
Hoop #1 Hoop #2

b bc2

l1 (l1/bc2 ≤ 2)
Vertical web
l2 (l2/bc2 ≤ 2) reinforcement
bc1

lbe

(b) SBE with two overlapping hoops and supplemental crossties

Figure 7 Examples of SBEs with (a) and without (b) overlapping hoops

In summary, this proposal addresses performance issues related to single perimeter hoops
with aspect ratio greater than approximately 2.0 and unsupported vertical web reinforcement
by: (1) limiting the length of a hoop leg along the perimeter of a boundary element to twice
the wall thickness, (2) requiring the use of 135-135 crossties for SBEs, where crossties are
used/allowed, and (3) requiring the use of 135-135 crossties or similar for web vertical
reinforcement over the height of the SBE.
References:

Segura, C. L., and Wallace, W. J., 2017a, “Seismic performance limitations and detailing of
slender RC walls,” ACI Structural Journal, 29pp, accepted for publication (minor revisions),
Oct 24, 2017.

Segura, C. L., and Wallace, W. J., 2017b, “Impact of geometry and detailing on drift capacity
of slender walls,” ACI Structural Journal, 32pp, accepted for publication (minor revisions),
Dec 1, 2017.

Welt, T.S. (2015). “Detailing for Compression in Reinforced Concrete Wall Boundary
Elements: Experiments, Simulations, and Design Recommendations,” Ph.D. Dissertation,
University of Illinois at Urbana-Champaign, 530 pp.

Arteta, C.A. (2015). “Seismic Response Assessment of Thin Boundary Element Specimens
of Special Concrete Shear Walls,” Ph.D. Dissertation, University of California, Berkeley,
240 pp.
ACI 318-19 CH 011f – Wall Boundary Detailing

Change Proposal:

Line # Change proposal material in strikeout/underline (code or commentary text,

figures, tables, etc.)
0 Changes to notation (Chapter 2)
hx = maximum center-to-center spacing of longitudinal bars laterally supported by corners of
crossties or hoop legs around the perimeter of the a column or a wall boundary element, in.
lbe = length of boundary element from compression face of member, in.
Changes to Section 18.7.5.2(a)
(a) Transverse reinforcement shall comprise either single or overlapping spirals, circular
hoops, or single or overlapping rectilinear hoops with or without crossties.
10 18.10.6.2(b) Where If special boundary elements are required by (a), the special boundary
element transverse reinforcement shall extend vertically above and below the critical section
at least the greater of lw and Mu/4Vu, except as permitted in 18.10.6.4(gh)

20 18.10.6.4 Where If special boundary elements are required by 18.10.6.2 or 18.10.6.3, (a)
through (hi) shall be satisfied:
...
(e) The boundary element transverse reinforcement shall satisfy 18.7.5.2 (a) through (ed)
and 18.7.5.3, except the value hx in 18.7.5.2 shall not exceed the lesser of 14 in. and two-
thirds of the boundary element thickness, and the transverse reinforcement spacing limit of
18.7.5.3(a) shall be one-third of the least dimension of the boundary element.
(f) Transverse reinforcement shall be arranged such that longitudinal bars or bundles of bars
around the perimeter of the boundary element shall be are laterally supported by a seismic
hook of a crosstie or corner of a hoop. leg such that The spacing between supported
longitudinal bars, hx, shall not exceed the lesser of 14 in. and two-thirds of the boundary
element thickness,. tThe length of a hoop leg shall not exceed two times the boundary
element thickness, and adjacent hoops shall overlap at least the lesser of 6 in. ches and two-
thirds the boundary element thickness.
ACI 318-19 CH 011f – Wall Boundary Detailing

30 (h) Web vertical reinforcement over the wall height For a distance above and below the
critical section specified in 18.10.6.2(b), web vertical reinforcement shall have lateral support
provided by the corner of a hoop or by a crosstie with seismic hooks at each end. Transverse
reinforcement shall have a vertical spacing not to exceed 12 in. and diameter of transverse
reinforcement shall satisfying 25.7.2.2.

(gi) Where the critical section occurs at the wall base, …

(hj) Horizontal reinforcement in the wall web…

40 R18.10.6.4 - Fifth paragraph:
Required transverse reinforcement at wall boundaries is based on column provisions.
Expression (a) … This considers that concrete spalling is likely to occur only on the exposed
faces of the confined boundary element. Tests (Thomsen and Wallace, 2004) show that
adequate performance can be achieved using vertical spacing greater than that permitted by
18.7.5.3(a). The limits on hx are intended to provide more uniform spacing of hoops and
crossties for thin walls. Tests (Thomsen and Wallace, 2004) show that adequate performance
can be achieved using vertical spacing greater than that permitted by 18.7.5.3(a).
50 Configuration requirements for boundary element transverse reinforcement and crossties for
web vertical reinforcement are summarized in Fig. R.18.10.6.4.1. A limit is placed on the
relative lengths of boundary element hoop legs because tests (Segura and Wallace, 2017;
Welt et al., 2015; Arteta 2015) show that a single perimeter hoops with supplemental multiple
crossties with that have alternating 90-degree and 135-degree hooks are not as effective as
overlapping hoops and crossties with seismic hooks at both ends if lbe exceeds approximately
2b.
60 These tests also show that loss of wall axial load carrying capacity of a walls can occur
immediately following damage to the wall boundary elements if web vertical reinforcement
within the plastic hinge region is not restrained with crossties with a seismic hook at each
end. Use of web crossties outside of boundary elements also results in a less abrupt transition
in transverse reinforcement used to provide concrete confinement and restrain buckling of
longitudinal reinforcement, which addresses potential increases in the neutral axis depth due
to shear (diagonal compression) and uncertainties in axial load.

Requirements for vertical extensions of boundary elements are summarized in Fig.

R.18.10.6.4.3 (Moehle et al., 2011).
70 Add new Fig. R18.10.6.4.1 as shown below and renumber current Fig. R18.10.6.4.1 to Fig.
ACI 318-19 CH 011f – Wall Boundary Detailing

R18.10.6.4.2, and current Fig. R18.10.6.4.2 to Fig. R18.10.6.4.3 (and change references to
these figures). Also revise Figure R18.10.6.4.2 to show crossties with seismic (135-degree)
hooks at both ends.
80 Add references:
Segura, C. L., and Wallace, W. J., 2017, “Seismic performance limitations and detailing of
slender RC walls,” ACI Structural Journal, 29pp, accepted for publication (minor
revisions), Oct 24, 2017.
Welt, TS, Massone LM, LaFave JM, Lehman DE, McCabe SL, Polanco P, “ Confinement
Behavior of Rectangular Reinforced Concrete Prisms Simulating Wall Boundary
Elements,” ASCE Journal of Structural Engineering April 2017 Volume 143.
Arteta, C.A. (2015). “Seismic Response Assessment of Thin Boundary Element Specimens
of Special Concrete Shear Walls,” Ph.D. Dissertation, University of California,
Berkeley, 240 pp.
90 Perimeter hoop Supplemental crossties Vertical web Through web
Crosstie
reinforcement

b bc2

bc1 (bc1 / bc2 ≤ 2) Horizontal web

reinforcement, Av

lbe

(a) Perimeter hoop with supplemental 135-degree crossties and 135-degree crossties
supporting distributed web longitudinal reinforcement.
Hoop Overlap Horizontal web Through web
Supplemental crosstie at least min of reinforcement, Av
(6 in. and 2b/3 Crosstie
Hoop #1 Hoop #2

b bc2

l1 (l1/bc2 ≤ 2)
Vertical web
l2 (l2/bc2 ≤ 2) reinforcement
bc1

lbe

(b) Overlapping hoops with supplemental 135-degree crossties and 135-degree crossties
supporting distributed web longitudinal reinforcement.
R18.10.6.4.1 Configurations of boundary transverse reinforcement and web crossties