You are on page 1of 15

ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 86-S6

Seismic Shear Strength of Circular Reinforced Concrete


Columns

by Ang Beng Ghee, M. J. N. Priestley, and T. Paulay

The need for a consistent seismic design philosophy to establish de- The performance of circular reinforced concrete col-
sign shear forces for reinforced concrete columns of bridges and umns under seismic attack has not always been exem-
buildings under seismic attack is discussed. It is shown that because
of a combination of the effects of conservatism in flexural strength plary, despite a common perception that they represent
design equations and approximations in representation of seismic the optimum shape for seismic resistance. The San Fer-
loads for analysis purposes, real shear forces generated during earth- nando earthquake of 1971 produced several examples
quakes may exceed three times the shear corresponding to code levels of flexural and shear failures of circular bridge col-
of seismic latera/loads. umns, 1 and similar failures have been reported from re-
Existing U.S. and New Zealand design expressions for shear
strength of circular columns are discussed and compared with results
cent 1apanese earthquakes. 2 During the 1967 Venezuela
from a comprehensive test program involving 25 circular columns earthquake, 3 the large circular columns of the Macuto
tested under axial load and cyclic lateral inelastic displacements. It is Sheraton Hotel suffered shear failures.
shown that existing design equations are inconsistent and very con- Although shear failures in these and other recent
servative for initial shear strength. The experimental results indicate, earthquakes have been common, the shear strength of
however, that the shear strength reduced with increasing flexural
ductility for displacement ductility factors in excess of 2. A model for
concrete columns under seismic load has not received as
the influence of flexural ductility on shear strength is advanced, with much emphasis as the flexural strength and ductility.
design equations for initial shear strength applicable for low flexural This can be attributed to three main causes:
ductilities, and for final shear strength applicable for fully ductile 1. Reasons for shear failures in the 1971 San Fer-
flexural designs.
nando and other recent earthquakes appeared rather
Keywords: columns (supports); ductility; dynamic loads; earthquake-resistant obvious and were particularly related to poor detailing,
structures; flexural strength; lateral pressure; reinforced concrete; shear such as the lapping of shear reinforcement in the cover
strength; structural design; tests.
concrete that spalled during seismic response.
Circular reinforced concrete columns have been fa- 2. By adopting conservative capacity design princi-
vored for bridge pier design for some time because of ples, it is a comparatively simple matter to insure
simplicity of construction and also because of omnidi- against shear failure, even if shear strength is not pre-
rectional strength characteristics under wind and seis- cisely known.
mic loads. Although less common in buildings because 3. Prediction of the design shear force requires an
of difficulties with detailing beam-column intersec- accurate knowledge of the flexural strength. Hence,
tions, and with integrating circular columns within determination of flexural strength of bridge columns
constraints imposed by glazing and curtain walls, they was seen as a more fundamental research problem than
are still sufficiently common to merit special consider- shear strength.
ation and appear to be currently enjoying an architec- Nevertheless, determination of the shear strength of
tural vogue. columns under seismic attack is a real problem. Until
recently, conservative detailing rules for ductility in
RESEARCH SIGNIFICANCE columns have generally resulted in requirements for
Research reported in this paper shows that ex1stmg confinement reinforcement being more critical than for
design equations in ACI 318 for shear strength of cir-
cular columns are conservative and inconsistent. De-
sign recommendations made in the paper would allow
Received Dec. 7, 1987, and reviewed under Institute publication policies.
a more rational and generally more economical design Copyright© 1989, American Concrete Institute. All rights reserved, including
for shear strength of circular columns subjected to seis- the making of copies unless permission is obtained from the copyright propri-
etors. Pertinent discussion will be published in the November-December 1989
mic actions. ACI Structural Journal if recieved by July I, 1989.

ACI Structural Journal I January-February 1989 45


Ang Beng Ghee, BE (Hons), ME, PhD (Canterbury), graduated from the Uni-
strength of axially loaded columns, and (4) the provi-
versity of Canterbury, Christchurch, New Zealand. He worked in Malaysia for sion by the designer of excess flexural reinforcement
two years after completing his doctorate. Presently, he is a structural engineer above the minimum amount necessary to satisfy the
with Morrison, Cooper and Partners in Wellington, New Zealand. His main
code level bending moment. For a ductile bridge design
interests are structural design and computer analysis, with particular emphasis
on earthquake engineering. based on lateral load levels established by dividing the
M. J. N. Priestley is Professor of Structural Engineering at the University of expected elastic response inertial loads by a reduction
California at San Diego, and has been extensively involved in research on the factor that may be as high as six or eight, it is the ac-
response of concrete structures. He is Chairman of A Cl Committee 53/, Con-
crete Masonry Structures; and a member of A CJ Committee 344, Circular Pre-
tual flexural strength that will be developed in the de-
stressed Concrete. He was previously on the faculty of the University of Can- sign earthquake rather than the nominal design level,
terbury, New Zealand, and is a past president of the New Zealand National even if this exceeds the design level by a factor of three
Society for Earthquake Engineering. In 1984, he and two other Canterbury
faculty received A Cl's Raymond C. Reese Structural Research A ward.
or more. It should be noted, however, that the conse-
T. Paulay, FACI, is a professor of civil engineering at the University of Can-
quence of extra flexural strength will be a reduction in
terbury, Christchurch, New Zealand, author of numerous papers on the design the level of ductility required during the design level
of reinforced concrete buildings for earthquake resistance, and coauthor of earthquake. Similar behavior will occur in columns of
Reinforced Concrete Structures. Dr. Paulay is a member of joint ACI-ASCE
Committee 445, Shear and Torsion; and ACI Committees 442, Response of
building frames designed for ductile seismic response.
Concrete Buildings to Lateral Forces; and 3/8, Standard Building Code. He is In both bridge and building designs, shear forces can
a past president of the New Zealand National Society for Earthquake Engi- be further enhanced by dynamic response effects and
neering. Prof. Paulay has twice been co-recipient of ACI's Raymond C. Reese
Structural Research A ward. variation in the position of contraflexure in the col-
umns from positions predicted by elastic structural
shear reinforcement- Now, with less conservative con- analyses based on code-specified distribution of lateral
finement equations being suggested for ductile columns loads.
[see Reference 4 for example], shear design is again be- The design level of column shear force V0 can thus
coming critical and deserves renewed and critical ap- be related to the shear force VCL corresponding to the
praisaL This paper discusses capacity design principles code-level lateral seismic load by the expression
necessary to establish a realistic level for the design
shear force for circular columns and presents results
from an experimental program designed to investigate (1)
the influence of transverse reinforcement, axial load,
and flexural ductility on the shear strength of circular where ¢ 1 is the flexural strength reduction factor; k 1 is
columns. a factor for flexural strength enhancement resulting
DESIGN SHEAR FORCE from material strengths exceeding nominal design lev-
Dependable ductile behavior from reinforced con- els; k 2 is a factor for conservatism in flexural strength
crete columns under seismic loading is dependent on equations; k 3 defines the excess flexural strength result-
developing flexural hinging at previously identified and ing from provision of excess reinforcement; and w,. is a
carefully detailed plastic hinges. It is essential that pre- shear amplification factor for dynamic response ef-
mature shear failure, either within the plastic hinges or fects.
elsewhere, does not occur, as shear failure, and in con- Current American design practice for spirally con-
junction with axial compression is nonductile and is as- fined building columns, as embodied in ACI 318-83
sociated with rapid strength, stiffness, and physical (revised 1986), 4 is to use a flexural strength reduction
degradation under cyclic loading. This was graphically factor of
demonstrated in the 1971 San Fernando earthquake and
in the recent Whittier Narrows earthquake of October
1987. (2)
To insure against shear failure, a capacity design ap-
proach must be adopted. That is, the shear capacity of
Thus the lower limit of ¢ 1 = 0. 75 applies for axial
the column is set higher than the shear corresponding to
load ratios Plj;Ag ~ 0.1.
development of maximum feasible flexural strength at
For bridge columns, a more conservative approach is
the designated plastic hinge locations.
Thus, for example, the design shear force for a suggested by A TC-6 5
bridge column is not directly related to the design hor-
izontal inertial loading specified by a code but instead cPJ = 0.9 - 2 (3)
depends on the maximum flexural strength that can be
developed in the plastic hinge regions of the column.
This maximum flexural strength will exceed the design The lower limit of ¢ 1 = 0.5 applies for axial load ra-
level for flexural strength because of (l) flexural tios Plf: Ag ~ 0.2.
strength reduction factors commonly incorporated in New Zealand design practice adopts c/>1 = 0.9 regard-
flexural design calculations, (2) material strengths less of axial load level when the potential plastic hinge
(concrete compression strength and reinforcement yield region is detailed for ductility. 6
strength) exceeding nominal design values, (3) the con- In New Zealand, k is generally taken as 1.15 (i.e., 15
1

servative nature of design equations for flexural percent enhancement possible due to over-strength ma-
46 ACI Structural Journal I January-February 1989
terials). In the United States, greater variation in rein-
forcement strength from the nominal value occurs, and ":E
'x2.4
a higher value of k 1 would be appropriate. However, in 0
E
the following discussions, k 1 = 1.15 will be assumed. :E_ 2.2
Reinforced concrete columns are generally designed 0
for flexure using the ACI compression stress block 4 (re- -§
a:: 2.0
gardless of section shape), an ultimate compression c(I)
strain of Ecu = 0.003 with no allowance for strain hard- E 1.8 n 4
ening of reinforcement. Extensive experimental and e3
c
0
theoretical studies of confined columns 7 have indicated ..c
c 1.6
that this approach is very conservative, particularly for lJJ

high levels of axial load P. This is because lateral pres- c(I) 1.4
sure from confinement reinforcement in the plastic E
0
hinge region causes a substantial increase in the :E
compression strength of the concrete, and the high sec- 1.2
tion strains at moderate to high ductility levels also re-
sult in higher steel strains and, hence, higher steel 1.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
stresses than are predicted based on an ultimate
Axial Load Ratio P/fcAg
compression strain of Ecu = 0.003. This can result in
some of the flexural reinforcement attaining stresses
higher than yield strength due to strain hardening. In Fig. 1-/nfluence of axial load on moment enhance-
Fig. 1, the ratio Mmax I M; is plotted against the axial ment ratio 7 (related to ACI flexural strength calcula-
tions with ¢ = 1. 0)
load ratio PI f:A~ for a wide range of column tests,
where Mmax is the maximum measured moment during
testing, and M; is the ACI ideal moment based on mea-
sured material strengths. 7 The degree of flexural
strength enhancement in Fig. 1 is clearly strongly de-
pendent on the axial load ratio and can be expressed as

4
(4a)

V0 3 r----t-------,{r-
p

' --;:---- > 0.1: VcL


J;Ag ~b)

Mmax
M;
r p
= 1.13 + 2.35 f.'Ag - 0 . 1jF

Experimental data fall within ± 15 percent of Eq.


(4). The appropriate level for k 2 would thus appear to
be the upper bound of the relationship given in Fig. 1,
corresponding to 1.15 times Eq. (4). OOL-~--~~~--~~-L--~--~~--
Simple expressions cannot be developed for k, and w, 0.2 0.4 0.6
in Eq. (1), as they will depend on design details. How- Axial Load Ratio f x P/fc Agl
ever, for regular building frames, suggestions for the
shear amplification factor w, are listed in the New Zea- Fig. 2-Designlcode shear ratio for ductile columns
land concrete design code NZS 3101: 1982. 6 designed for flexure according to ACI 3184
Fig. 2 plots the ratio V0 1 VCL implied by Eq. (1),
based on the previous values for ¢ 1, k 1, and k 2 , assum- suit in economics of flexural design, but would also
ing k 3 and w, both = 1.00 (i.e., no excess reinforce- clearly result in reduced shear design levels.
ment or shear amplification). It will be seen that very
large ratios of V0 1 VCL would need to be adopted to in- COMPARISON OF U.S. AND NEW ZEALAND
sure against shear failure, particularly if the very con- CODES
servative flexural strength reduction factors of A TC-6 Code shear strengths of circular columns
are adopted. Also plotted in Fig. 2 is the constant value Current U.S.' and New Zealand 6 design practices
of V 0 1 VeL = 1.47 that results from flexural design consider a portion of the design shear force to be car-
adopting Eq. (4) as a predictor of flexural strength with ried by concrete shear resisting mechanisms V (the
¢1 = 0. 9 independent of axial load as recently pro- concrete contribution), with the remainder Vs carried by
posed by Priestley and Park. 7 This would not only re- truss mechanisms involving transverse reinforcement
ACI Structural Journal I January-February 1989 47
steel ratio, taken as 0.5 AjAg where A,, is the total
0 •7 0 .,/,f)()mm
d;;: 0.80
~ = JOOkNm
longitudinal steel area, and A, is the gross column area
for a circular column, and
o.s
P(4h - d)
Mm = Mu- 8 (9)

where M" is the ultimate moment occurring in conjunc-


tion with V", and h is the section depth (h = D for cir-
cular sections). An upper limit of

(a) L1mifs fo Concr~f~ Confr~bvf1on (b) Compartson lor Typn;ol Circular Colvmn v, = 0.29 .jJ: · .JI + 0.29?1 A" (10)

Fig. 3-Comparison between U.S. and New Zealand limits the concrete contribution given by Eq. (8).
code provisions for contribution of concrete to shear The seismic provision of Appendix A of Reference 4
resistance (stresses in MPa; -1 MPa = 145 psi; 1 mm implies that for ductile columns, v, shall be assumed to
= 0.03937 in.) be zero if Plf/A, ~ 0.05, though the relevant clauses
are ambiguous.
and 45 deg concrete compression diagonals. Thus
NZS 3101
In regions of columns outside plastic hinges, the
v; (5)
concrete contribution is taken as

ACI 3I8 requires a strength reduction factor of cPs = (II)


0.85 to be used in Eq. (5), while NZS 3IOI 6 allows cPs=
1.00 to be used when V0 is established by the capacity where the basic concrete shear stress vb is given by
design approach developed earlier in this paper. As ex-
0.08.jJ: ~ vb = (0.07 + lOpw) JJ: ~ 0.2 JJ: (12)
perimental results presented in this paper will be as-
sessed by comparison with U.S. and New Zealand code
Eq. (II) and (I2) were recommended by the joint
provisions, a summary of the relevant requirements for
ACI-ASCE Committee 426 in 1979, but were not
members loaded with axial compression follows.
adopted by ACI 3I8.
In plastic hinge regions for axial load levels of P ~
Shear carried by concrete O.I J:' Ag, v, is taken as zero because the primary contri-
In both countries, the concrete contribution to shear butions to v, are likely to be undependable when wide
strength is t::~ ken as
flexural cracks occur under ductile flexural response.
For axial compression loads of P ~ O.I J:' Ag, v, is
(6) taken as

where v, is the nominal shear stress carried by the con- (13)


crete; bw is the width or diameter of the column; and d
is the effective depth of the section, taken as 0.8 times where vb is given by Eq. (12). Eq. (I3) presumes that the
the diameter D of the column. shear capacity of the concrete shear resisting mecha-
nisms increase with axial load because of reduced crack
ACI 318 widths (and hence enhanced aggregate interlock) and
The simplified expression given by Eq. (ll-4) of Ref- improved compression shear transfer in the more ex-
erence 4 is generally adopted by designers. This equa- tensive compression zone. Eq. (13) approaches the
tion is expressed as follows in Eq. (7), in S.l. units value given by Eq. (II) when PIJ:'Ag ~ 0.4.
A comparison between the U.S. and N.Z. provisions
v, = O.I67(l + 0.0725P!Ag)JJ: (7) for concrete contribution to shear resistance is pre-
sented in Fig. 3. In Fig. 3(a), the ACI 3I8 "maxima"
where vn f:, and the axial stress PI A, are in MPa. The [Eq. (10)] and "minima" [Eq. (7)] are compared with
designer has the option of adopting the more detailed, NZS maxima [Eq. (II) or (13) with vb = 0.2 JJ:l and
and hence presumably more exact and less conserva- NZS minima [Eq. (II) or {13) with vb = 0.08 Jl:].
tive, Eq. (ll-6) expressed in Eq. (8) in S.l. units Provisions for both ductile and nonductile columns are
shown. From these curves, it will be seen that much
17'1 V,,d higher concrete contributions to shear strength are per-
v, = 0.158 .Jf: + I7.2p.-- (8)
M"' mitted by the ACI 318 provisions. However, this does
not necessarily imply that higher values will result from
where V" is the shear due to factored loads at the sec- the ACI provisions in any particular case. Fig. 3(b)
tion under consideration; Pw is the longitudinal tension compares the U.S. and N .z. provisions for a typical
48 ACI Structural Journal I January-February 1989
400 mm (16 in.) diameter column with a total steel ra-
tio of 0.03 (hence p., = 0.015) and a shear span-to-ef-
fective depth ratio of aid = 2.5. Comparison of the 1agonal Shear
Crack
ACI [approximate-Eq. (7)-and exact-Eq. (8) and
(9)1 methods reveals apparentl"Y anomalous behavior.
The exact method is more conservative over most of the
axial load range than the approximate equation and is
hyperbolic in nature, predicting a 17 percent increase in
v, as the axial load ratio increases from 0 to 0.3, but a
160 percent increase as the load ratio increases from 0.4 !'
to 0.48.
It will also be noted from Fig. 3(b) that for this ex-
ample, the nonductile provisions of the N.Z. code are
less conservative than ACI 318, and that the ductile
provisions of NZS 3101 are less conservative than the
ACI exact and approximate equations for PIJ:' Ag
greater than 0.16 and 0.19, respectively.

Shear carried by transverse reinforcement


Both U.S. and N.Z. concrete codes recommend the
calculation of shear carried by truss mechanisms in-
volving transverse reinforcement in circular columns by
representing the section as an equivalent rectangle of
width D and effective depth d = 0.8D. This approxi-
mation is unnecessary, however, as computation of the
shear-resisting forces in the hoops exposed by a pre- (b) Plan
sumed 45 deg diagonal tension crack and resolving par-
allel to the applied shear force, as shown in Fig. 4, re-
Fig. 4-Shear carried by transverse reinforcement for
sults in the expression circular column

(14)

where Ash is the area of the hoop or spiral bar of yield


strength !vh with pitch sand diameter D'. In Eq. (14),
it is assumed that all transverse reinforcement exposed
by the crack is at yield, and that the pitch s is suffi-
ciently small compared with the diameter D' to vali-
date the integral averaging implied by Eq. (14). For low
( 11 vanes as lisf«J m Table f)
D' Is ratios, Eq. (14) can be up to 10 percent noncon-
{a J Elevaf1on (b) End View {C J X· Secf1an
servative.8

TESTING OF CIRCULAR COLUMNS FOR SHEAR Fig. 5-Column unit dimensions (1 mm = 0.03937 in.)
STRENGTH
Despite the prevalence of circular columns in prac- Test unit details
tice, there have been surprisingly few tests to establish Because of the dearth of information relevant to the
shear strength. Farodji and Diaz de Cassio (cited in seismic shear strength of circular columns, a series of
Reference 9) investigated the shear strength of twenty- 25 circular columns of 400-mm (15.75-in.) diameter,
one 250 mm (10 in.) diameter circular columns. How- considered to be approximately one-third scale models
ever, only four columns contained web reinforcement, of typical bridge columns, were constructed and tested
and none were subjected to cyclic load reversals typical under cyclic reversals of lateral loading as part of a
of seismic loading. The results of these tests are cited by major investigation into the strength and ductility of
ACI-ASCE Committee 4269 in support of current ACI bridge columns. 8 Variables in the test program included
provisions. Khalifa and Collins 10 tested five circular axial load level, longitudinal reinforcement ratio,
columns with circular hoops, under double bending. transverse reinforcement ratio, and distribution and as-
Four of their columns were tested with monotonic pect ratio. The column units were tested as simple ver-
loading, while one column was tested with load re- tical cantilevers. As shown in Fig. 5, the columns were
versals. They found that shear strength was typically 20 constructed monolithically with a column base an-
percent higher than predicted by ACI equations, and chored to a structural test floor.
that compression field theory 11 gave a more accurate Longitudinal reinforcement was deformed steel of 16
prediction. or 24-mm (0.63 or 0.945-in.) diameter, and nominal
ACI Structural Journal I January-February 1989 49
Table 1 - Column unit details and material strengths
Aspect Longitudinal reinforcement Transverse reinforcement
ratio, J:.
Unit MIVD PIJ:'A, MPa Quantity p, f,,, MPa db, mm s, mm p, X !0' J,,, MPa
I 2.0 0 37.5 20-HD16 0.0320 436 6 60 5.094 328
2 2.0 0 37.2 20-Dl6 0.0320 296 6 60 5.094 328
3 2.5 0 36.0 20-HD16 0.0320 436 6 60 5.094 328
4 2.0 0 30.6 20-HD!6 0.0320 436 !0 165 5.146 316
5 2.0 0 31.1 20-HD16 0.0320 436 6 40 7.642 328
6 1.5 0 30.1 20-HD16 0.0320 436 6 60 5.094 328
7 2.0 0 29.5 20-HD16 0.0320 448 6 80 3.82! 372
8 2.0 0.2 28.7 20-HD16 0.0320 448 6 30 !O.Jg9 372
9 2.5 0.2 29.9 20-HD16 0.0320 448 6 30 10.189 372
- -f - - - -
!0 2.0 0.2 3!.2 20-HD16 0.0320 448 !2 120 !0.189 332
II 2.0 0.2 29.9 20-HD16 0.0320 448 6 60 5.094 372
!2 1.5 0.1 28.6 20-HD16 0.0320 436 6 30 10.189 328
13 2.0 0.1 36.2 20-HD16 0.0320 436 6 30 10.!89 326
14 2.0 0 33.7 9-HD24 0.0324 424 6 60 5.094 326
15 2.0 0 34.8 12-HD!6 0.0!92 436 6 60 5.094 326
16 2.0 0.1 33.4 20-HD16 0.0320 436 6 60 5.094 326
!7 2.5 0.1 34.3 20-HD16 0.0320 436 6 60 5.094 326
18 !.5 0.1 35.0 20-HD!6 0.0320 436 6 60 5.094 326
19 1.5 0.1 34.4 20-HD!6 0.0320 436 6 80 3.821 326
20 !. 75 0.175 36.7 20-HD!6 0.0320 482 6 80 3.821 326
21 2.0 0 33.2 20-HD!6 0.0320 436 6 80 3.821 326
22 2.0 0 30.9 20-HD!6 0.0320 436 10 220 3.859 310
23 2.0 0 32.3 20-HD16 0.0320 436 12 160 7.642 332
24 2.0 0 33.1 20-HD16 0.0320 436 10 110 7.719 310
25 1.5 0 32.8 20-Dl6 0.0320 296 - - - -
016 = 16 mm-Grade 275 deformed bar; HDI6 = 16-mm Grade 380 deformed bar; HD24 = 24-mm Grade 380 de-
formed bar.
Plain round bars such as R6 were used.
I MPa = 145 psi; I mm = 0.03937 in.

yield strength of either 275 or 380 MPa (40 or 55 ksi). monitored by dial gages mounted on steel rods passing
All bars had a 90-deg hook at the base, and the top through, and anchored to, the core concrete. Full de-
ends were welded to an annular steel plate to insure ad- tails of the instrumentation are given in Reference 8.
equate anchorage. Plain round bars were used for
transverse spiral reinforcement. Laps of transverse re- Load application and test procedure
inforcement were effected by full strength single flare The simulated lateral seismic load was applied by a
welds. 500 kN (112 kip) double-acting servo-hydraulic actua-
Concrete specified for the columns had a maximum tor, connected to a steel loading collar around the top
aggregate size of 20 mm (0.79 in.), a slump of 75 mm of the column. A I 00 ton single acting jack was used to
(3 in.), and a target 28 day strength of 30 MPa (4350 provide the required constant axial compression to col-
psi). At the start of each column test, standard con- umn units subjected to axial load.
crete cylinders were tested to establish compressive The loading sequence represented by Fig. 6 was
strength at the time of column testing. Table I includes applied to all columns. The sequence started with five
details of the column dimensions, axial load ratios, initial cycles of horizontal load to a load level at 75
longitudinal and transverse reinforcement, and mate- percent of the lateral load V; 1 required to impose the
rial strengths. theoretical ideal flexural strength, based on the ACI
compression stress block, measured material strengths
Instrumentation and an ultimate compression strain of 0.003. As shown
Transverse reinforcement strains were monitored in Fig. 6(a), the first full cycle to this load level was
with 5 mm (0.20 in.) electric resistance strain gages. used to define the yield displacement by extrapolating
Lateral displacement at the level of load application straight lines from the origin through the peak load
was measured by a 200 mm (7 .87 in.) travel linear po- displacement points at 0. 75 V; 1 to the theoretical
tentiometer. In addition to these basic measurements, strength V;1. The average of the values in both positive
curvatures within the potential plastic hinge region were and negative loading directions was taken as the yield
50 ACI Structural Journal I January-February 1989
displacement, and the displacement ductility factor was =!-
~ 1.1.: l':.mcut
defined as the ratio of actual displacement to this yield ~ B~----6y ------------~
displacement.
Subsequent loading was by five complete displace-
ment controlled cycles to successive displacement duc-
tility factors of p. = 1.5, 2, 4, 6, and 8, as shown in Fig.
6(b), unless premature shear failure caused an early
termination of testing.

RESULTS OF TEST PROGRAM


(a J Experimen,af De fin" ion lbJ o,spfacemenl Sequence
Except for Column 9, all column units had insuffi- of Yield Displacement

cient margins of ideal shear strength over the ideal


flexural strength to satisfy the capacity design require-
Fig. 6-Sequence of imposed displacements for column
units
ments presented earlier in this paper. Consequently,
shear failures were expected from most of the tests, graded. A side view of Unit 9 after testing is shown in
though in many cases considerable flexural ductility Fig. 8(a).
was anticipated. Fig. 7(c), (d), and (e) show hysteresis of three units,
On the basis of the experimental results, the columns with different levels of axial load, from the moderate-
were classified into four catagories, as follows ductile category. In these and other units of this cate-
gory, the ideal flexural strength was attained or
Ductile flexural (D·F) exceeded at displacement ductility levels of p. ~ 4. In
Column units that achieved ductility levels of p. ~ 6 further loading to displace the units beyond p. = 4,
without any indication of shear failure. sudden degradation in strength and stiffness occurred.
Yield of the spiral reinforcement at the sides of the
Moderately ductile with shear failure (MD·S) units caused excessive widening of diagonal cracks with
Column units that achieved ductility levels of 4 < p. a corresponding loss of aggregate interlock capacity.
~ 6 before exhibiting shear failure. The reduction in the strength of the concrete shear-re-
sisting mechanisms caused a demand for further con-
Limited ductile with shear failure (LD·S) tribution from spiral reinforcement to carry the applied
Column units that achieved ductility levels of 2 < p. shear, which could not be provided. Coupled with this,
~ 4 before exhibiting shear failures. high strain was also imposed on the spirals as they were
mobilized to confine the dilating concrete. In some
Brittle shear failure (B·S) cases, especially for units with significant axial
Column units that exhibited shear failure at p. ~ 2. compression [e.g., Unit 8, Fig. 7(e)], the combined ef-
Generally these units were not able to sustain shears fects resulted in fracture of the spirals at the sides of the
corresponding to the ideal flexural strength. unit despite spiral ultimate tensile strains in excess of 20
percent.
Load-displacement hysteresis loops Comparison of the loops of Fig. 7(c), (d), and (e),
Fig. 7 shows hysteresis loops for typical examples of show that flexural strength enhancement ( vm.J V;f) in-
the four categories, and conditions of typical column creased with axial load leveJ,1 as indicated in Fig. 1.
units at the end of the test are illustrated in Fig. 8. A There is also less pinching of the stable flexural loops
summary of the results obtained from each column test, when higher axial load is applied, but degradation of
including the failure category, is included in Table 2. the loops after initiation of shear failure is more rapid
Only one column (Unit 9) produced a fully ductile at higher load levels.
response. Fig. 7(a) and (b) show hysteresis loops for the Fig. 8(b) shows a side view of Unit 13 at the end of
early and latter stages of testing of this unit. In this and testing. Note the increased region of cover spalling
the other hysteresis loops of Fig. 7, the lateral load Vis compared with Unit 9 [Fig. 8(a)], the fractured spiral
given in kN and also as the ratio VI V:>. where V,1 is the reinforcement, and the dilation of the core indicated by
shear force corresponding to the ideal flexural strength the barrelling of the longitudinal reinforcement.
Mil. Fig. 7(f) and (g) compare loops for limited ductile (2
Unit 9 developed stable hysteresis loops at flexural < p. ~ 4) column units with Plf'Ag = 0 and 0.1, re-
ductility levels of p. ~ 6, with peak loads exceeding V,1 spectively. No unit with Plf;Ag = 0.2 fell within this
by a considerable margin. At p. = 8, excessive yield of group. Behavior is similar to the columns with moder-
the transverse reinforcement within the plastic hinge ate ductility but with shear failure occurring rather ear-
region was noted. This occurred at opposite ends of the lier. In each case, stable flexural hysteresis loops with
loaded diagonal rather than at the sides and was a re- peak lateral loads exceeding Vif formed at p. ~ 2.
sult of the confinement capacity of the spiral being ex- For the LD-S columns, initial diagonal cracking gen-
ceeded. The yielding spiral reinforcement allowed the erally developed at an early stage with a crack inclina-
longitudinal compression reinforcement to develop in- tion to the column axis between 40 and 50 deg. At later
cipient buckling, and the lateral load capacity de- stages of testing, a pair of wide, intersection diagonal
ACI Structural Journal I January-February 1989 51
500 500
6

ill
IJ.::
II.~ 2
I I
375
_j_
250
~ T,.
~
-40 40

V,:: 28ii<N -L
,.11 :; B.Jmm

!UNIT 9!
I I I
.] -2 J,l.: -8 -6
-500
(a) Ductile, IJH (b) Ductile, 6SIJS8

v;, .:: 363kN


V, 1 = 201.kN .:31 "B.Smm
Ll, = 6.6mm
-I.
,.,r- - !UNIT 131
I I
!UNIT 751 J,l.:: -6 -4

(c) Moderately Ductile, P (d) Moderately Ductile, P 0.1


f(:Ag = O f(:Ag

jl. =2
""" 2
500
I I
,_~ 375
I

-40

V, 1 :: 3D4kN
V, 1 "' 356kN ll, = s.t.mm
..1," 6.8mm
1.0 IUNIT1t.l
.J J.s-375
"" J _, .J J
(e) Moderately Ductile, f~A = 0.2 (f) Limited Ductile, _P____
c g f{,Ag - 0

Fig. 7(a)-(f)-Hysteresis loops for different modes of failure (1 kN = 225 lb; 1 mm


= 0.03937 in.)
cracks eventually formed, usually at a smaller angle to common. Fig. 8(d) shows a typical column (Unit 16) at
the column axis than the original diagonal cracking. the end of testing, illustrating the brittle failure mode
Fig. 8(c) shows the condition of Unit 14 at the end of and the steep crack inclination.
test with strong diagonal failure planes apparent.
Fig. 7(h), (i), and (j) show hysteresis loops for units
Strength of concrete shear resisting mechanisms
exhibiting a brittle shear failure mode. Typically,
(a) At onset of diagonal cracking-As the so-called
though not exclusively, these units were unable to sus-
concrete contribution to shear strength in a number of
tain the ideal flexural strength although the shortfall in
code equations is taken as the load at the onset of di-
capacity was never more than 20 percent. Shear fail-
agonal cracking, it is of some interest to compare ex-
ures generally occurred after maximum strength was
perimental values with code equations listed earlier in
attained at a ductility of about J1- = 1.5. As is apparent
this paper.
from the hysteresis loops, strength and stiffness degra-
To allow comparison of results with code equations,
dation for these units was particularly severe. The sud-
the shear stress at diagonal cracking vcr has been calcu-
den drop in lateral load-carrying capacity from the
lated as
maximum load corresponded to excessive widening of
diagonal cracks that opened up suddenly and extended
almost from corner to corner. Fracture of spiral rein-
vcr = (15)
forcement at comparatively low displacements was 0.8Ag
52 ACI Structural Journal I January-February 1989
IVV,,~~ttf
400 1.0----

300

60 -20

V,," 281 kN
.1, = 9.0mm -200 V,, = 1.02kN
_L_
I .1y = 5-5mm
""LR;-;- I -300

~ ~-5 -4 -3 -2 -!5
p
~-;--~-::~ -400 1.0
(g) Limited Ductile,
f(:Ag = 0.1 (h) Brittle p
f(:Ag = O

froclure

~~~:ure
of spuol

ofsp1ro1

V., = l.68kN V.t = t.77kN


..1, = 6.,mm ..11 :: 7.8mm

~-· -375
.-===-
~~.l JJ., -
(i) Brittle, P
T'A = 0.1
(j) Brittle, _P___ = 0 . 175
c g f(:Ag

Fig. 7(g)-(j)-Hysteresis loops for different modes of failure (1 kN = 225 lb; 1 mm


= 0.03937 in.)

where A 8 is the gross section area, and 0.8Ag is taken as


the effective shear area. This area approximately cor-
responded to the area of the confined concrete core and
was considered a more realistic assessment of effective
area than current code equations, which would use
0.8D 2 • Note that this value exceeds Ag.
To further allow comparison between column units
with different concrete strengths, shear stress is ex-
pressed in the form vj JJ:. vcr and fi are expressed in
MPa, and to obtain vj JJ: in psi units, it is necessary (a) Ductile (b) Moderately Duct.1le

to multiply by Jf45 = 12.04.


Experimental data are plotted against axial load ra-
tio P!f;Ag in Fig. 9(a). Columns with different aspect
ratios MIVD are identified by different symbols. Note
that shear at diagonal cracking increased with axial
load ratio and with decreasing aspect ratio. Separating
the data into units with the same aspect ratio indicated
linear increase of shear at diagonal cracking for each
aspect ratio, as indicated by the dashed lines in Fig.
9(a), with a significant difference between units of as-
pect ratio 1.5 and 2.0 but very little difference between
(c) ... mited Ductile (d) Brittle
units of aspect ratio 2.0 and 2.5.
Also plotted in Fig. 9(a) are the predictions from the Fig. 8-Condition of units with different failure modes
nonductile code provisions presented earlier in this pa-
per. The ACI minimum [Eq. (7)] and exact [Eq. (8), simplified Eq. (7), is apparent. Eq. (ll ), which repre-
based on the maximum p"' used in the tests], is very sents the upper bound to the N.Z. design equation for
conservative and, again, the illogical aspect of the ACI V 0 is also very conservative, but ACI Eq. (10), which is
exact approach, in that it provides values less than the intended only as an upper bound limiting Eq. (8), gives
ACI Structural Journal I January-February 1989 53
Table 2 - Strengths of column units
Unit details Theoretical strengths Observed behavior
M p
Unit - - M,t,• v,f,' V,' V,,• (}, ** Vm,tt V""
VD J:A, p,. p, kNm kN kN kN deg kN kN ll" Mode•••
I 2.0 0 0.0320 0.0051 245 306 88 !53 50 320 232 2.5 LD-S
2 2.0 0 0.0320 0.0051 176 220 88 177 38 228 140 4.0 MD-S
3 2.5 0 0.0320 0.0051 242 242 88 !51 45 298 210 4.0 LD-S
4 2.0 0 0.0320 0.0051 239 295 85 !50 48 295 210 1.4 B-S
5 2.0 0 0.0320 0.0076 241 301 131 !50 40 340 209 2.4 LD-S
6 1.5 0 0.0320 0.0051 241 402 88 177 45 390 302 1.3 B-S
7 2.0 0 0.0320 0.0038 234 293 74 163 45 280 206 1.6 B-S
8 2.0 0.2 0.0320 0.0102 285 356 198 252 40 475 277 4.0 MD-S
9 2.5 0.2 0.0320 0.0102 281 281 198 244 45 385 187 7.8 D-F
10 2.0 0.2 0.0320 0.0102 288 360 176 261 43 450 274 4.0 MD-S
II 2.0 0.2 0.0320 0.0051 281 351 100 240 47 404 304 2.5 LD-S
12 1.5 0.1 0.0320 0.0102 262 437 175 325 47 527 352 3.0 LD-S
13 2.0 0.1 0.0320 0.0102 290 363 174 210 43 443 269 4.0 MD-S
14 2.0 0 0.0324 0.0051 243 304 86 159 46 311 225 2.0 LD-S
15 2.0 0 0.0192 0.0051 163 204 86 134 42 230 144 4.0 MD-S
16 2.0 0.1 0.0320 0.0051 276 345 86 219 42 379 293 1.5 B-S
17 2.5 0.1 0.0320 0.0051 281 281 86 203 37 329 243 2.0 LD-S
18 1.5 0.1 0.0320 0.0051 277 462 86 290 38 507 421 1.4 B-S
19 1.5 0.1 0.0320 0.0038 281 468 65 273 42 436 371 1.3 B-S
20 1.75 0.175 0.0320 0.0038 334 477 65 327 37 487 422 1.5 B-S
21 2.0 0 0.0320 0.0038 250 313 65 157 45 258 193 1.1 B-S
22 2.0 0 0.0320 0.0039 235 294 62 !59 38 280 218 1.5 B-S
23 2.0 0 0.0320 0.0076 236 295 124 193 45 339 215 2.0 LD-S
24 2.0 0 0.0320 0.0077 244 305 125 198 42 338 213 4.0 LD-S
25 1.5 0 0.0320 - 178 297 - 180 50 233 233 1.2 B-S
• Ideal flexural capacity.
' Shear corresponding to Mif = eV,1.
'Shear carried by 45-deg truss [Eq. (5)].
1 Shear at onset of diagonal cracking.
•• Angle of diagonal crack to column axis.
" Maximum shear force sustained.
" vm- v,.
••• D-F = ductile flexural; MD-S = moderately ductile, shear; LO-S = limited ductile, shear; B-S = brittle, shear.

Data. M/VO 0.7 0


/
0 1.5 M /0

----.
0
(> 1.75
a.6 vo""",.,/ /
0.6 • 2.0 \• _..._...

~cr
0.5
D 2.5
~cu
TrJ 0.5 /
/
/

. -- --
_... _...
_...
ACI Max
I Eqn.10)

0.4 0.4
Yrl
NZS 3101 NZS 3101
0.3 !Eqn.tt)
0.3 (Eqn.11)

ACI Min !Eqn 7) ACI Min (Eqn. 7)

_ _ :=·~~:=:.==JACI"Exact"
Jj_ __ , / (EqnB J
~=;..:;11===~=
Ei ·--· M
,__ '_,)·5--
s--· 1 1ACl "Exact"
!Eqn 8)

VO o.tL-----~------~~v~o___ J_ _ _ _ __L~-...-
0.1 L----~----~------~----~~~--- o 0.05 o.to 0.15 0.20
0 0.05 0.15 0.20a. to
p ;rcAg P/f(:Ag
Ia) At Onset of Diagonal Cracking (b) At Maximum Lateral Load

Fig. 9-Shear carried by concrete shear resisting mechanisms compared with code equations (v" r; in MPa; 1 MPa
= 145 psi)
54 ACI Structural Journal I January-February 1989
a reasonable estimate of average v,, values obtained
from the tests. vcu =0.37a.(1+3P/f(;AgJ...fif
(b) At maximum shear load-U.S. and N.Z. codes 0.6
assume that the concrete shear contribution remains
unchanged as the total shear increases, with the bal- 0.5
ance being carried by a 45 deg truss mechanism. It is
thus of interest to compare the concrete contribution to
1 vc 0.4
shear at onset of diagonal cracking v,, with that at
O:..[jj:
maximum lateral load v,,. A conservative estimate of v,, c Q3
was obtained by subtracting the shear carried by truss
mechanism [using Eq. (14)] from the total shear and Vcr = 0.28a. ( 1+3P/f/:AgJ...ff[
Q2
dividing by the effective shear area of 0.8Ag. Only col-
umns with brittle shear (B-S) or limited ductile (LD-S) f2s;;JJQ: • At diagonal crocking
0.1
behavior were included in this assessment, as the inclu- o At maximum load
sion of results of ductile or moderate ductile columns
where flexure governed initial response would provide 0.05 0.10 0.15 Q20
unrealistically low values for v,", since identical col- P/f(;Ag
umns with unlimited flexural strength would presum-
ably have attained higher shear strengths. Fig. 10-Shear carried by concrete shear resisting
Results are plotted in Fig. 9(b), with data points mechanisms compared with proposed equations (v" C'
identified by aspect ratio, and plotted against axial load in MPa; 1 MPa = 145 psi)
ratio. The trends apparent in Fig. 9(a) are repeated in
Fig. 9(b) though the concrete contribution to shear at
maximum lateral is approximatelv 30 percent higher 0.8
than at the ons~t of diagonal cracking. The code equa-
tion [Eq. (7) through (10)], also plotted on Fig. 9(b),
thus display even greater conservatism when compared 0.6
with shear at maximum lateral load.
The experimental data suggest a strong dependence 0.1.
on aspect ratio for M I VD < 2, but a less definite rela-
tionship for larger aspect ratios. The data may thus be
further nondimensionalized by expressing it in the form 0.2
vJ(cx.Jl:) where
2
ex= - - - ~ 1.0 (16)
(MIVD)
Ps (%/
Data at onset of diagonal cracking and at maximum
lateral load in this form are plotted against axial load Fig. 11-Influence of transverse reinforced ratio on ul-
ratio as solid and hollow data points, respectively, in timate shear strength when P If; = 0 (v,, f; in MPa; 1
Fig. 10. Also plotted are best-fit straight lines through MPa = 145 psi)
the data. At the onset of diagonal cracking
sumption of a truss mechanism with a 45 deg diagonal
v,, = 0.28cx(l + 3PifiAg) .JJ: (17) compression strut and with all spiral reinforcement at
agrees with the experimental data with an average ex- yield. Although the visual evidence supported the as-
periment/theory ratio of 1.00 and a coefficient of vari- sumption of approximately 45 deg diagonal cracking in
ation of 8. 7 percent. the initial elastic and low-ductility phases of testing, it
At maximum lateral load is of interest to examine the data for more explicit con-
firmation.
v," = 0.37cx(l + 3Pif/Ag) .JJ: (18) In Fig. II, the nondimensionalized shear stress at
maximum lateral load vj JJ: = V"/(0.8Ag.jJ[) is plot-
agrees with the experimental data with an average ex-
ted against the spiral reinforcement content Ps for nine
periment/theory ratio of 1.00 and a coefficient of vari-
column units having MIVD = 2.0 and without axial
ation of 7.0 percent.
load, where the volumetric spiral reinforcement ratio is
It will be noted that Eq. (17) and (18) differ from the
upper limit of Eq. (11) only in the value of the leading
numerical coefficient, which is 40 and 85 percent higher Ps = 4A,h7rD' = 4Ash (19)
for Eq. (17) and (18), respectively. 7r(D') 2s D's

Influence of shear reinforcement on performance All nine column units in Fig. 11 failed in shear with
The results of Fig. 9 and 10 relating to concrete shear limited or no ductility. The data appear to vary linearly
contribution at maximum load were based on the as- with the spiral reinforcement content, and the best-fit
ACI Structural Journal I January-February 1989 55
Influence of ductility on components of shear
resistance
1, 2 The degradation of shear strength with ductility is
examined further in Fig. 13, which shows plots of shear
1,0 strength versus cumulative ductility factors for three
columns typifying limited ductile, moderate ductile, and
ductile response. Cumulative ductility is defined as the
o.a sum of all positive and negative ductility peaks occur-
:L
V;f ring prior to a given stage of testing. Thus five cycles
0.6 to p. = ± 1.5 contributes a cumulative ductility of~~' =
15.
0.4 In Fig. 13, the solid line and data points represent
total applied shear force. The dashed line represents the
0.2
shear carried by spiral reinforcement, using the expres-
sion

(21)

Fig. 12-Degradation of shear strength with ductility


where (} is the measured crack inclination to the longi-
line gives the relationship tudinal axis, and j,h is the spiral reinforcement stress
indicated by strain gages. It will be noted that when j,h
= fyh and(} = 45 deg, Eq. (21) simplifies to Eq. (14).
0.38 + 28p, (20) The difference between the solid and dashed lines in
Fig. 13 represents the concrete contribution. In all three
The coefficient for p, in Eq. (20) agrees to within 3 plots, the concrete contribution remains essentially
percent with analytical expressions based on a 45 deg constant after initial development of ideal flexural
truss mechanism and average material strengths for the strength until a cumulative ductility factor of ~~' ::::: 35
corresponds to the end of cycling to p. = 2. Subsequent
test units. 8
The influence of spiral reinforcement on response to this, the concrete contribution gradually decreases
can further be examined by comparing the envelope until the maximum load is achieved, with much more
force/displacement curves of columns with different p, rapid degradation of strength in the postultimate stage
values but identical in all other respects. Fig. 12 com- of testing. For Unit I, with limited ductile response,
pares envelopes in terms of VI V, 1 versus p. plots for maximum load corresponds to p. = 2, and rapid
three columns with aspect ratio M I VD = I. 5 and axial strength degradation occurs at this stage. Unit 9, with
load ratio PIf.' Ag = 0.1. Envelope curves are given for ductile response, shows an approximately linear de-
the first and fifth cycles at each ductility level so that crease of Vc with cumulative ductility up to the stage
the degree of degradation occurring at a fixed ductility when buckling of longitudinal reinforcement caused the
can be calculated. Unit 19 failed before a fifth cycle at test to be terminated.
p. = 4 could be applied, so a lower envelope for the
third cycle is plotted. DESIGN RECOMMENDATIONS
Unit 19, with p, = 0.0038, contained insufficient The data presented in Fig. 12 and 13, and contained
shear strength to enable the ideal flexural strength V,1 to for other columns in the complete report on this study, 8
be attained. Maximum load occurred at p. = 1.5, and suggest that a model for shear strength degrading with
the subsequent strength degradation was very rapid. increasing flexural ductility could be developed. A pos-
Unit 18, with p, = 0.0051, enabled the ideal flexural sible form of this relationship, which is similar to one
strength to be reached, but insufficient shear strength recently proposed by the Applied Technology Council
existed to allow to develop normal flexural strength en- for retrofitting highway bridges, 12 is shown in Fig. 14.
hancement due to concrete confinement and steel strain An initial shear strength V, is assumed to apply for dis-
hardening. Maximum lateral load was also achieved at placement ductilities of p. :::;; 2. At higher ductilities, the
p. = 1.5, but subsequent strength degradation was less shear strength degrades until a final value V1 is attained
severe than with Unit 19. At p, = 0.0102, the spiral re- when the flexural ductility capacity p.1 is reached. Meth-
inforcement content of Unit 12 was sufficient to allow ods for estimating p.1 are presented elsewhere. 7 •13
significant flexural strength enhancement to develop, If the shear force corresponding to development of
and shear failure was deferred until p. = 3. Postpeak the real flexural strength is less than V1, as shown in
strength degradation was comparatively gradual. Fig. 14 by Line 1, then full flexural response is assured.
The behavior represented in Fig. 12 was typical of Line 2 corresponds to the case when the shear corre-
other subsets of the data with different axial load ra- sponding to real flexural strength exceeds V1 but is less
tios and aspect ratios from the values for Fig. 12. 8 than V;. The ductility JJ.c achieved is defined by the in-
56 ACI Structural Journal I January-February 1989
tersection of Line 2 with the shear strength degradation
curve. Line 3 represents the case when shear force cor- 300 ---Vs 1.00
responding to real flexural strength exceeds V,. A brit- -v 0.75
tle failure will occur at p. :::; 2. 200 X.
0.50 Vit
100
Initial shear strength 0.25
The results from the tests reported herein are ade- .
0
~
quate to provide an estimate of the initial shear strength
0
{a) Limited Ductility
V; of circular columns. As previously noted, initial ~
shear strength could be calculated using the additive 0
principle, with a 45 deg analogous truss mechanism. c...
Combining Eq. (14) and (18) ~ 4fXJ
l:J
§ 300

i.e., V, = 0.37a (1 + 3P!J:' Ag) (22)


71' D'
../J[' Ae + lAsh fvlz-----;-

Final shear strength


The data indicated that as the flexural ductility fac-
tor increased and the concrete contribution to shear de-
creased, the angle (} of the analogous truss mechanism
to the longitudinal axis decreased, increasing the shear
carried by the analogous truss, as indicated by Eq. (21).
It also appeared that the total shear strength decreased
approximately linearly with increasing flexural ductil-
ity, as indicated in Fig. 14.
The angle (} can be estimated from a lower bound
plastic theory solution 4
(c) Ductile
tanO = f (23) Fig. 13-lnf/uence of cumulative ductility factor on
~ components of shear resistance (1 kN = 225 /b)

where

(24) A
-...
and v < 1 is a factor for reduced effective compression <:
~
strength of the diagonal compression strut. Vecchio and '--
::::::,..
Collins 14 have explained v as being a consequence of re-
duced compression strength of concrete in the presence ClJ
of lateral tensile strain. It is probable that under con- ~
ditions of cyclic reversals of inelastic displacement
causing intersecting diagonal cracks, values for v will be
u:
lower than for the monotonic cases primarily consid- 'CJ
ClJ
ered by Vecchio and Collins. For the columns tested in ..c:
this project, a value of v = 0.2 was found to be appro- V)
priate at a ductility of p. = 6. A limit to 0 is imposed
by a corner-to-corner failure plane, and Thurlimann 15
has suggested a lower design limit of(} = 25 deg. 0 1.0 2.0 IJ.c iJ. f
Insufficient data were available from the tests to de-
fine the degraded concrete contribution with any con- Displacement Ductility
fidence. In particular, the influence of axial load was
Capacity/ iJ.
not clear, but appeared to be less significant than was
the case for initial shear strength. Based on the limited Fig. 14-Re/ationship between shear strength and dis-
data, a tentative value of placement ductility capacity
ACI Structural Journal I January-February 1989 57
v,1 = 0.185 JJ: (25) tility, and the inclination to the longitudinal axis of the
diagonal compression struts of the truss mechanism de-
i.e., 50 percent of the basic concrete contribution at creased. Tentative proposals for degraded shear
zero axial load is suggested, regardless of axial load strength, and the form of the shear strength versus
level. However, because the spiral reinforcement pro- flexural ductility relationship were developed but re~
vided a secondary function of confining the core con- quire verification by further testing.
crete, and thus, apparently reduced the severity of deg- More testing is also required to establish the shear
radation of the concrete contribution, Eq. (25) should strength of circular columns at axial load ratios outside
be reduced in proportion toPs when p, ::::; 0.01. the comparatively narrow range investigated in this
Combining the effects of the concrete contribution program. Other factors that may affect the results, in-
and truss mechanism results in cluding biaxial lateral response and the influence of
double bending on shear strength, need investigation.
(26)
ACKNOWLEDGMENTS
where The experimental results in this pap..:r formed part of Ang's PhD
studies, supervised by Priestley and Paulay. Financial assistance of
the Road Research Unit of the New Zealand Roads Board and of the
V,! = 18.5ps Jl:·A . : : ; 0.185 JJ:·A, (27)
University of Canterbury are gratefully acknowledged.

and
NOTATION
A, effective shear area = 0.8A,
A, gross section area
A,, area of hoop or spiral bar
A, total longitudinal steel bar area
b, effective column width = D
noting that cot(25 deg) = 2.15. D gross column diameter
D' diameter of confined core
CONCLUSIONS d effective depth of section
Capacity design procedures are needed to define the 1: cylinder compressive strength of concrete, MPa
f., stress in hoop or spiral reinforcement
design shear force for bridge and building columns. J,., yield stress of hoop or spiral reinforcement
These procedures must take into account the very high L shearspan
shear amplification that can result from flexural M moment
strength enhancement of confined columns designed in M, column flexural strength based on ACI design equation
accordance with conservative flexural strength equa- Mmu = measured flexural strength
P axial load on column
tions. s spacing of hoop or spiral reinforcement along longitudinal
Examination of the ACl 318 design expressions for axis
shear strength of concrete columns indicates anoma- V shear force
lous behavior for the concrete contribution Vc. The ex- V, shear force carried by concrete shear~resisting mechanism
V,1 V, at high values of p.
act equation [ACl 318 Eq. (11-6)] is more conservative
V,, V, at low values of p.
over most of the axial load range than the simplified VeL shear force corresponding to code loads
expression [ACI 318 Eq. (11-4)] and is hyperbolic in V:., shear at onset of diagonal cracking
nature with little influence of axial load on Vc until V0 design shear force
about Plj; Ag = 0.4, after which Vc increase extremely V1 final shear strength at high p.
rapidly. V: initial shear strength at low p.
V:1 shear corresponding to ideal flexural strength
Results from a series of 25 circular columns tested as V, shear force carried by truss mechanism involving transverse
vertical cantilevers under axial load and cyclic reversals reinforcement
of inelastic lateral displacements indicated that the V,, V, at high values of p.
shear strength was dependent on the axial load level, v, basic concrete shear stress in NZS 310 I'
v, nominal shear stress carried by concrete
the column aspect ratio (MIVD), the amount of trans-
v,1 nominal shear stress corresponding to V,,
verse spiral reinforcement, and the flexural displace- v,., nominal shear stress corresponding to V:.,
ment ductility factor. At low flexural ductilities, the v" nominal shear stress corresponding to V:,
additive principle, based on a concrete contribution v,, nominal concrete shear stress at maximum lateral load
plus a 45 deg truss mechanism involving the spiral re- v, nominal shear stress of truss mechanism
v, nominal shear stress at maximum lateral load
inforcement and diagonal concrete compression struts,
a aspect ratio factor, defined by Eq. (16)
described behavior well. However, existing U.S. and 1/1 mechanical reinforcement ratio, defined by Eq. (26)
New Zealand design equations for the concrete contri- </>1 flexural strength reduction factor
bution were found to be very conservative. Based on </>, shear strength reduction factor
the test results, suggested design equations were ad- p. displacement ductility factor
factor for reduced effective compression strength of diago-
vanced.
nal strut
At flexural displacement ductilities of p. > 2, the p, volumetric ratio of hoop or spiral reinforcement to core vol~
shear strength degraded gradually with increasing due- ume
58 ACI Structural Journal I January-February 1989
p,. longitudinal tension steel ratio Department of Civil Engineering, University of Canterbury, Chri-
{) measured crack inclination to longitudinal axis stchurch, July 1985, 408 pp.
w, dynamic shear amplification factor 9. ACI-ASCE Committee 426. "Shear Strength of Reinforced
Concrete Members," Proceedings, ASCE, V. 99, ST6, June 1973, pp.
REFERENCES 1091-1187.
I. Jennings, P. C., Editor, "Engineering Features of the San Fer- 10. Khalifa, 1. U., and Collins, M.P., "Circular Reinforced Con-
nando Earthquake of February 9, 1911,"Report No. EERL 71-02, crete Members Subjected to Shear," Publication No. 81-08, Depart-
California Institute of Technology, Pasadena, 1971, 512 pp. ment of Civil Engineering, University of Toronto, Dec. 1981, 103 pp.
2. Ohashi, M., et al., "Overview of the State of Practice in Earth- II. Collins, Michael P., and Mitchell, Denis, "Shear and Torsion
quake Resistance Design of Highway Bridges," Earthquake Resis- Design of Prestressed and Non-prestressed Concrete Beams," Jour-
tance of Highway Bridges, ATC-6-1. Applied Technology Council, nal, Prestressed Concrete Institute, V. 25, No.5, Sept.-Oct. 1980, pp.
Palo Alto, 1979, pp. 43-66. 32-100.
3.Hanson R. D., and Degenkolb, H. J. "The Venezuela Earth- 12. "Seismic Retrofitting Guidelines for Highway Bridges," ATC-
quake July 29, 1967," American Iron and Steel Institute, New York, 12/FHW A/RD-83/007, Applied Technology Council/Federal High-
1969, l76pp. way Administration, Washington, D.C., 1983, 205 pp.
4. ACI Committee 318, "Building Code Requirements for Rein- 13. Zahn, F. A.; Park, R.; Priestley, M. J. N.; and Chapman, H.
forced Concrete (ACI 318-83) (Revised 1986)," American Concrete E., "Development of Design Procedures for the Flexural Strength
Institute, Detroit, 1986, Ill pp. and Ductility of Reinforced Concrete Bridge Columns," Bulletin,
5. Seismic Design Guidelines for Highway Bridges, ATC-6, Ap- New Zealand National Society for Earthquake Engineering (Welling-
plied Technology Council, Berkeley, 1981, 199 pp. ton), V. 19, No.3, Sept. 1986, pp. 200-212.
6. "Code of Practice for the Design of Concrete Structures," (NZS 14. Vecchio, F., and Collins, M. P ., "Stress-Strain Characteristics
3101: 1982), Standards Association of New Zealand, Wellington, of Reinforced Concrete in Pure Shear," Proceedings, IABSE Collo-
1982, 127 pp. quium (Delft, 1981), International Association for Bridge and Struc-
7. Priestley, M. 1. N., and Park, R., "Strength and Ductility of tural Engineering Ziirich, pp. 211-225.
Concrete Bridge Columns Under Seismic Loading," ACI Structural 15. Thiirlimann, Bruno, "Shear Strength of Reinforced and Pre-
Journal, V. 84, No. I, Jan.-Feb. 1987, pp. 61-76. stressed Concrete-CEB Approach," Concrete Design: U.S. and Eu-
8. Ang, B. G.; Priestley, M. J. N.; and Paulay, T., "Seismic Shear ropean Praclices, SP-59, American Concrete Institute, Detroit, 1979,
Strength of Circular Bridge Columns," Research Report No. 85-5, pp. 93-115.

ACI Structural Journal I January-February 1989 59

You might also like