ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 89-51

Seismic Resistance of Type 2 Exterior Beam-Column

Joints Reinforced with Inclined Bars

by A. G. Tsonos, I. A. Tegos, and G. Gr. Penelis

An experimental investigation of the behavior of external beam-col- of all internationally known researchers in the field ap-
umn joints with inclined reinforcing bars under seismic conditions is pears.
presented. A simple technique to prevent these elements from failing
in premature, explosive cleavage shear fracture was implemented for Paulay, Park, and Phillips5 first introduced the idea
the first time. Twenty full-scale reinforced concrete exterior beam- of using inclined main reinforcement to prevent brittle
column subassemblies were tested. The primary variables were the failure in short reinforced concrete coupling beams.
amount of inclined bars, the ratio of the column-to-beam flexural Following the work of Paulay, Park, and Phillips, 5
capacity, and the joint shear stress. Test results showed that use of Minami and Wakabayashi6 and Tegos and Penelis' ap-
crossed inclined bars in the joint region is one of the most effective
ways to improve the seismic resistance of exterior reinforced concrete plied the idea in short columns and carried out some
beam-column joints. experiments. The results showed that the columns with
inclined reinforcing bars performed considerably better
than those with conventional reinforcement. Beam-col-
Keywords: beam-column frame; beams (supports); columns (supports); con· umn joints have many similarities in geometry, state of
nections; cyclic loads; earthquake-resistant structures, fiber reinforced con-
crete; hinges (structural); joints (junctions); metal fibers; reinforced concrete; stress, and mechanical behavior with short columns and
reinforcing steel; shear strength; structural analysis. coupling beams. For these reasons, the efficiency of a
new nonconventional reinforcing pattern by the use of
crossed inclined bars is examined as a way of improv-
Current design philosophy permits moment-resisting ing the seismic performance of Type 2 exterior beam-
frames subjected to seismic loading, apart from a few column connections. 8•9
exceptions, to be induced into the inelastic range where
the forces that develop in parts of the structure will ex-
ceed their design values. 1•3 In this phase of inelastic in- RESEARCH SIGNIFICANCE
tensity, the beam-column joints are obliged to resist This study investigates the improvement in the seis-
high horizontal and vertical shear stresses coming from mic behavior of exterior reinforced concrete beam-
the adjacent beams and columns. This occurs during a column joints resulting from the presence of inclined
large number of inelastic cycles and while the joints reinforcing bars in their core region. The tests on exter-
need to dissipate large energy amounts. However, in nal beam-column joints reported herein show that it is
spite of systematic research that has been carried out possible to achieve highly ductile behavior by reinforc-
investigating the behavior of reinforced concrete beam- ing these structural elements with the proposed new,
to-column connections in the past 25 years in the nonconventional reinforcing pattern.
United States, New Zealand, and Japan, we have not The comparison of this new reinforcing type with the
succeeded in improving satisfactorily this behavior, the other well-known nonconventional type of using steel
specific features of which are extreme decrease of (a) fibers showed that the former pattern has remarkable
strength; (b) stiffness; and (c) energy dissipation capac- advantages that offer better seismic beam-column joint
ity, all of which may lead to collapse of the whole performance, economy in construction, and structural
structure. simplicity of design.
Attempts at improving beam-column joint perform-
ance resulted in nonconventional ways of reinforcing, ACI_Structural Journal, V. 89, No. I, January-February 1992.
Rec~1ved Apr. 16, 199!, and reviewed under Institute publication policies.
such as the use of fiber concrete. For this type of rein- Copyng~t © 1992, Amencan Concrete Institute. All rights reserved, including
fo~ement, more research is proposed by ACI-ASCE the makmg of copies unless permission is obtained from the copyright propri-
etors. Pertinent discussion will be published in the November-December 1992
Committee 352, 4 in which the recent design philosophy ACI Structural Journal if received by July I, 1992.

ACI Structural Journal I January-February 1992 3

A.G. Tsonos is a researcher of reinforced concrete structures, Department of
conventionally reinforced in the joint region (Type S
Structural Engineering, Aristotle University, Thessaloniki, Greece. He received specimen) and one identical specimen reinforced with
his PhD from that university. Dr. Tsonos is the author of 10 t~chnical papers. inclined bars and hoop reinforcement in this region
His research is. primarily concerned with the inelastic behavior of reinforced
concrete structures, structural design, and fiber reinforced concrete.
(Type X specimen). The reinforcing bars in the beams
and columns were the same in both types of specimens.
I. A. Tegos is an assistant professor of reinforced concrete structures, Depart- The specimens of each series differed only in reinforce-
ment of Structural Engineering, Aristotle University. He received his PhD from
Aristotle University. Dr. Tegos has authored 40 papers concerning the behav-
ment details at the joint region. The second specimen of
ior of reinforced concrete beams and columns in bending and shear under seis- each series was reinforced with four crossed, inclined
mic conditions, fiber reinforced in torsion, and concrete technology. bars bent diagonally across the joint core, as shown in
G. G. Penelis is a professor in the Department of Structural Engineering, Ar-
Fig. 1, instead of the four intermediate longitudinal
istotle University. Dr. Penelis is also head of the department's laboratory of bars in the column of the first conventionally rein-
reinforced concrete structures. His main research and design interests concern forced specimen (Fig. 1). These four intermediate bars
reinforced concrete structures.
are the vertical joint shear reinforcement of these spec-
imens, as proposed by codes/·3 and their purpose is to
TESTING PROGRAM sustain the vertical joint shear forces. A development
Specimens and variables length for deformed bars in tension 1 is the extension of
The experimental program included 20 specimens in both inclined reinforcing bars and vertical shear rein-
eight series as detailed in Table 1. All specimens have forcements past the joint region (Fig. 1).
the same dimensions, as shown in Fig. 1. The sixth series consisted of three specimens. The
Each of the first five series consisted of two exterior first two, s6 and x6, were the same as the previously
reinforced concrete beam-column subassemblies, one described series' specimens. The third specimen, S~. was
Table 1 - Summary of experimental program
Joint Vertical Inclined Concrete
Series Specimen Co 1umn Beam transverse joint shear reinforcing compressive 1dh* hb/co I umn
reinforcement reinforcement reinforcement reinforcement bars strength (mm) bar diameter MR v** pt%
(psi)

51 3~14 [g 3$14 2$141012 $!4 3$8 4$14 0 5360 340(225) ~1.43(20) 2.79(1.40) 7 .36( 12) 0.75(0.47)
I
XI " " 3$8 0 4$14 3770 340 ( 268) 21.43(20) 2. 79(1.40) 8. 77(12) 0. 75(0 .47)

52 2$10 IDJz$10 2~12101 2~12 3$8 4$14 0 3770 340(230) 30 .00(20) I. 38( I. 40) 9.16( 12) 0. 75(0.47)
l$!0 l$10
II
xz " " 3$8 0 4~14 3480 340(239) 30.00(20) 1.42(1.40) 9. 54(12) 0. 75(0.47)

s3 2~10 Ia 2$10 2~12101 2~12

2~10 2~10
3$8 4$14 0 2750 340(269) 30.00(20) 1.04(1.40) 13 .22( 12) 0.75(0.47)
III
x3 " " 3$8 0 4~14 3910 340(226) 30.00(20) 1.10(1.40) 11.09(12) 0.75(0.47)

IV
54 3~10 o 3$10 2~14
2$12
IOI 2$12
2~14 3$8 4$!4 0 3040 340(299) 30.00(20) 0.97(1.40) 17 .58(12) 0. 75(0.47)

x4 " 3~8 0 4$14 2460 340(332) 30.00(20) 0.94(1.40) 19.54(12) 0.75(0.47)

2~14 !Ol2~14 3$8 4$14 340(274) 21.43(20) 1.20(1.40) 17 .90(12) 0.75(0.47)

55 4$14 1014$14 0 3620
1~10 l$10
v
x5 " " 3$8 0 4$14 3190 340(292) 21.43(20) 1.17(1.40) 19.08(12) 0. 75(0.47)

56 2$14 [OJ 2$14 4$141014$14 3$8 4~14 0 4780 340( 238) 21.43(20) 1.04(1.40) 15.58(12) 0.75(0.47)

VI x6 " " 3$8 0 4~14 3910 340(263) 21.43(20) 1.06(1.40) 17 .22(12) 0.75(0.47)

5. 4$1410]4$14 " 3~8 0 0 4200 340(254 I 21.43(20) 1. 79(1.40) 16.62(12) 0. 75(0.47)

6

x7 2$14 !DJ2$14 3$8 0 6~10 2610 340(322) 21.43(20) 1.06(1.40) 21.09(12) 0. 75(0.47)
VII
x8 " 3$8 0 4~10 2750 340(314) 21.43(20) 1.03(1.40) 20.53(12) 0. 75(0.47)

PI " " 0 0 ~ 2320 340(342) 21.43(20) 1.03(1.40) 22.38(12) 0(0 .47)

y1 " " 0 0 0 3330 340( 285\ 21.43(20) 1.02( 1.40) 18.66(12) 0(0.47)

VI II F1 " 0 0 0 2460 340(332) 21.43(20) 1.04(1.40) 21. 71(12) 0(0.47)

01 " 3$8 0 0 2900 340(306) 21.43(20) 1.04(1.40) 20 .00( 12) 0. 75(0.47)

F2 " " 3$8 0 0 3480 340(279) 21.43(20) 1.00(1.40) 18.27(12) 0. 75 ( 0. 4 7)

NOTE: c/>8,q,IO,c/>12,c/>14 = bar with diameter 8 mm, 10 mm, 12 mm, 14 mm.

I psi = 0.0069 MPa; I mm = 0.039 in.
Numbers outside the parentheses are the provided values.
Numbers inside the parentheses are required by ACI-ASCE Committee 352.
•t .. = <if, (psi)d,/75 .JJ: (psi)
**oy = shear stress as a multiple of .J1[ psi.
Summary of specimens' steel yield stress, in ksi, bar size: c/>8= 71.74, c/>10=67.45, c/>12 = 76.67, c/>14 = 70.30.

ACI Structural Journal I January-February 1992

 A

TOMH A-A
TOMH B-B
-y~t4tt4
hc•200
.L 3 4tl4

1-hc•200-1

Fig. 1 - Typical specimens of experimental program (dimensions in mm, 1 mm =

0.039 in., ld = development length for deformed bars in tension)

also conventionally reinforced, the only difference be- a. The percentage of inclined reinforcing bars.
ing that its joint vertical shear reinforcing bars were b. The ratio of column moment strengths versus
placed in the other columns' opposite sides, thus in- beam-moment strength at the joint, hereafter called
creasing the moment strength of the columns (see Ta- MR .
ble 1). .c. The horizontal joint shear stress determined by the
All these series were used to determine the efficiency parameter 'Y• as is defined by ACI-ASCE Committee
of the inclined bars when they replace an
equal amount 352. 4
of vertical joint shear reinforcement. Specimen S~ was Committee 3524 specifies maximum allowable joint
used to answer the possible question that the improve- shear stresses in the form of 'YJ!I, where joint shear
ment obtained from the proposed inclined reinforcing stress factor vis a function of the joint type (i.e., inte-
bars is simply caused by an increase in the columns' rior, exterior, etc.) and of the severity of the loading,
moment strength due to the presence of the inclined and f: is the concrete compressive strength. Lower lim-
bars' development lengths (Fig. 1). its of flexural strength ratio MR are also confirmed by
The seventh series consisted of two specimens rein- the Committee. Thus for the beam-column connections
forced in the joint region with different percentage of examined in this investigation, the lower limits of MR
inclined bars compared with that of specimen X 6 • The and 'Yare 1.40 and 12, respectively.
influence of the increase in the amount of inclined re- Exterior beam-column specimens with flexural
inforcing bars in the seismic behavior of the exterior strength ratio greater than 1.40 and joint shear stresses
beam-column joints is examined with the aid of this se- lower than 12./1: psi are considered to be the specimens
ries of specimens. with optimal MR and 'Y values. On the contrary, exte-
The characteristics of eighth series specimens in the rior beam-column specimens with flexural strength ra-
beam-column connection region were the following. tio less than 1.40 and joint, shear stresses higher than
Plain concrete and concrete from the same batch rein- 12./1: psi are considered to be the specimens with non-
forced with 1.0 percent by volume steel fibers were used optimal MR and 'Y values. The first series specimens
in the joint region of specimens P 1 and F 1, respectively, have optimal values for the parameters MR and v val-
while higher strength concrete was used in the joint re- ues. The second series specimens had joint shear stress
gion of specimens 0 1, Ytt and F2• Adequate shear rein- much lower than 12./1: psi, while their flexural strength
forcement consisting of stirrups was provided in the ratios were very close to 1.40. All other specimens used
beam-column connections of specimens 0 1 and F 2 , have nonoptimal values for MR and 'Y. The choice of
while specimens P 1, Y 1, and F 1 had no beam-column these values was made intentionally to see clearly the
connection reinforcement. The eighth series specimens seismic behavior improvement of specimens reinforced
had the same beam and column reinforcement as the with inclined bars, while the seismic behavior of their
sixth series specimens S6 and X 6 and were used to ex- corresponding conventionally reinforced specimens
amine the efficiency of the proposed reinforcement would be ineffective. All specimens had the same con-
pattern compared with the use of steel fibers as rein- fining reinforcement except for the unreinforced speci-
forcement.10-12 Details concerning reinforcements and mens P 1 and Ytt and the same development length for
material properties are summarized in Table 1. the hooked beam bars. 4 Due to the changes in material
The principal variables of the testing program were: properties and physical constraints, the _experimental
ACI Structural Journal I January-February 1992 5
 r- 700 --+~--- 2.800 ----t-700 1 Table 2 - Comparison of theoretical and experi-
mental results
I
one-way actuators M, Mbmax
8 a,
" Series Specimen MRcal MRexp vru/ M,y M,y kN
.. v""

0
0
,.,.
N
1
0
0
~
I

II
S,
X,
s,
2.79
2.79
1.38
2.17
1.71
1.42
7.36
8.77
9.16
6.88
9.53
1.25
1.40
8.59 1.15
1.30
1.54
1.15
23.42
26.52
5.62
X, 1.42 1.26 9.54 8.41 1.19 1.25 17.24
S, 1.04 1.07 13.22 9.57 0.95 1.04 8.43
III
X, 1.10 1.02 11.09 9.00 1.04 1.15 13.53
S, 0.97 1.31 17.58 10.18 0.74 0.78 7.36
IV
axial load jack X, 0.94 1.18 19.54 12.18 0.77 0.84 14.06
S, 1.20 1.41 17.90 10.94 0.78 0.84 12.44
4,200
v
X, 1.17 1.27 19.08 13.70 0.82 0.94 17.63
s, 1.04 1.27 15.58 9.85 0.78 0.79 7.15
Fig. 2- Test setup (dimensions in mm, 1 mm = 0.039 VI X, 1.06 0.98 17.22 14.05 0.99 1.10 16.67
in.)
s; 1.79 1.95 16.62 11.28 1.01 1.03 8.55
X, 1.06 1.18 21.09 15.26 0.80 0.93 13.05
VII
X, 1.03 1.26 20.53 12.90 0.72 0.83 11.26
P, 1.03 1.82 22.38 9.50 0.57 ,0.57 4.37
Y, 1.02 1.53 18.66 8.93 0.68 0.68 3.92
VIII F, 1.04 1.32 21.71 13.10 0.80 0.80 5.18
0, 1.04 1.65 20.00 9.80 0.59 0.63 5.63
F, 1.00 1.41 18.27 9.60 0.70 0.70 12.60

= design values of parameters

= actual values of parameters
= beam flexural strength
= maximum beam bending moment during the first cycle of
loading
= maximum beam bending moment durng the test

Fig. 3 - General view of the experiment

a = hysteresis loop pinching
I kN = 225lb
values of MR and 'Y are different from the design val-
ues. The design and experimental values of MR and 'Y
The loading of the specimens S 1, X 1, P" Y" F 1, S6 ,
for each specimen are listed in Table 2.
and X 6 consisted of several load reversals in the inelas-
tic range. The maximum displacement was increased
Test setup and instrumentation for each cycle after the first cycle of loading. All the
The tests of T -shape specimens were carried out in other specimens were loaded according to the load his-
the testing frame shown in Fig. 2. The column inflec- tory shown in Fig. 4.
tion points are simulated by the aid of specific arrange- The axial load was imposed by a hydraulic jack on
ments that are jointed by the testing frame and with the one of the two specimen columns. In this column, the
free ends of the specimen columns with hinges. Thus axial load was kept constant at approximately 0.40Pb
the horizontal and vertical displacement of the col- during the test. Depending on the magnitude of the ax-
umns' ends are restrained while they can rotate freely. ialload imposed on the beam's shear forces, the axial
Due to the manner in which the specimens were at- load of the other column changed continuously.
tached to the test frame, no magnification of the loads
due to P-1:. effects was introduced. Lateral loading was TEST RESULTS
applied to the beam's end by two one-way actuators. Plots of the applied load versus the displacement of
The load applied by the actuators was measured with the load point for representative specimens of this pro-
two load cells attached to the specimens. The load point gram are shown in Fig. 5 and 6.
displacement was measured by a specific potensiome- The specimens' stiffness was reduced with increasing
ter. The general view of the experiment is shown in Fig. cycles of loading. Stiffness decrease was smaller in the
3. specimens with inclined bars (Type X), compared with
6 ACI Structural Journal I January-February 1992
that of their conventionally reinforced counterparts (of 6(+)mm
65
Type S) and especially when the displacement was near 60 @)
55
zero, as can be seen from comparison of the speci- so
45
mens' hysteresis loop pinching (Table 2). Specimens of 40
®
3!j,
Type X exhibited higher strength and energy dissipa- 30
25
tion capacity in relation to those of their equivalent

~fI
20
15
specimens of Type S. The eighth series specimens
showed extreme decrease of strength, stiffness, and en- 0

1~
ergy dissipation capacity.
I5
For the sake of easier observation of strength degra- 20
dation, diagrams of percent yield strength versus the 25
30
displacement ductility are plotted (Fig. 7). The yield 35
40 ®
was measured during the test from the plot of the ap- 45
so
plied load versus the displacement of the load point and 55
60 cycle number
@)
corresponded to the case when significant decrease in 65
6(-)mm
the stiffness of the subassemblage occurred. From these
diagrams the following observations can be made:
Specimens S~o XI, x2, x4, Xs. x6, x7, and Xs performed
very well by maintaining their yield strength for dis- Fig. 4 - Loading sequence

Vo

170KN vo
60
1/ 160 K
I / II
1~0
./ v / I
so
'r- 2/
I/ 141

jo I jll 140 7 II I 6
30
v
1'0
20
'IV 1/ I20 1/111
/ I
-10 -90-111-70-60-50-40-30-20-10 ~ II
1 I -10 -90-60-70-60-50-<10-30-20 -10 ~ ~v I I
)/ t::: ~~ I I 1/
• vv b P;~ 1/ I J A

1/ I II l vvv vv 1/
I
1/1/ I I
'II 1/~ / /
1/ II J a f,o 20 JO 40 so so 10 eo omm 1/ I ~ ~ fto 2010 40 607080901 !lamm
I 1/V
v I 111 I
v 20

-30
I 1/ v ~v v VI I
VI
-20

-30
I I II/~ 1/ l/11/ I
v l11
"0
I 5 4 v v 1/1--'
{ 2
-40
I
vv./ -SO

-60
I
-50

-·~
v I specimen 51
-70
specimen x 1

VD

~~OK
1
40
l c

I3 I IY 7ls_
9

I ~H
~4-~~~·mM~~~HY

-7p-&>-5HO-l0-~-10 u 20
I
f7
r1 fT

~.~~ w uV
10
-7 -60-50-40-30-20-10 V-~v A

1/r;,r;, ~~1020 30 ..0~60 Omm

/rf'f 'I ~~r ~ ao•o so~ mm

fJ-1o

-20
I
~ 21
"'~rr
I
f++-t-bf'YJ,flbf./f.jm-ft-30 -t--+--1-+-+---1
I -3o
I/ I
f-1, ~'*fH.P.'!-V_,j_J-·b -t--+--1-+-+---1 7 -40

97J--Is3'
--~spec imen s2
specimen x 2

Fig. 5 -Load-versus-deflection response for Specimens XI, Sl, x2, s2

ACI Structural Journal I January-February 1992 7
 Fig. 6- Load-versus-deflection response for Specimens X6J S6J s;, F 2

140
o specimen s,
140 o specimen P,
c spec::iml!n 5 2
"specimen Y,
11 specimen S3
.:. specimen F,
• specimens.,. • specimen 0 1
• specimen S11
... specimen Se
• specimen F2
... specimen 5 8
" specimen 5$
X SpeCimen 56
G specimenX 8

10 12

Displacement ductility

10
Fig. 7(a) -Maximum load carried by Specimens S1
through s~ at various displacement levels. Displacement ductility

Fig. 7(c) - Maximum load carried by Specimens P 1,

Y~> F1, 0 1, F2, S6J s;, X 6 at various displacement levels

..
~

m
c the beam near the column. At the conclusion of the
.., o specimen x, test, their joints and critical column regions near the
•
~ 80
a specimen XI!
1>. specimen x3
• specimen X.o~
joint were almost intact, while the damage concen-
• specimen X11
... specimen X11 trated in their plastic hinge regions and had two char-
" specimen X7
60
G specimen X8 acteristic features: concrete rupture in the beam com-
pression zones and buckling of beam reinforcement
10 12 13
(Fig. 8).
Displacement ductility
The second failure mode is that of Specimen S1 •
There was again formation of a plastic hinge in the
Fig. 7(b)- Maximum load carried by Specimens X 1 beam near the column and more damage concentration
through X 8 at various displacement levels
in this region, but there was little damage in the joint,
placement ductilities of 5 or higher. Specimens with in- with partial loss of joint concrete cover (Fig. 8).
clined bars had a higher load-carrying capacity than the The third failure type is that of Specimens S2 , S3, X3 ,
conventionally reinforced specimens. For S2 , S3 , S4 , S5 , and ~. All specimens developed flexural hinges in their
S6 , S~, and for all the eighth series specimens, the load- beams. Damage occurred both in this region and on the
carrying capacity was sharply reduced after a displace- beam-column joint (Fig. 8).
ment ductility of approximately 2. In the fourth failure type, which is that of Specimens
X 5, X7 , and S6 , despite the appearance of hairline cracks
Failure modes during the first two cycles of loading, both in the joint
Five distinct failure modes were observed: and at the end of the beam, subsequent cycles resulted
The first is the failure mode of Specimens X 1 and X2• in an increase of the width of the cracks only in the
Both specimens developed flexural hinges at the end of joint (Fig. 8).
8 ACI Structural Journal I January-February 1992
 specimen xl specimen sl specimen x2 specimen s2

specimen x6 specimen s6 specimen S6 specimen F2

Fig. 8- Cracking configuration of specimens Xi> Sl> X 2, S2, X 6, S6> s;, F2 • (Cracking configuration of Specimens
S3, X 3 is identical with that of Specimens S2, X 6 ; cracking configuration of Specimens X 5, X 7 is identical with that of
Specimens;; and cracking configuration of Specimens S4, S5, X 4, X 8, P 1, Y 1, 0 1, F 1 is identical with that of Speci-
mens S6> F2 )

The fifth failure mode is that of Specimens S4 , S5, S6 , in the pastic hinge region (Fig. 8). However, for Speci-
X4 , X8 , and of all the eighth series specimens. In this men s2, the load-carying capacity deteriorated after the
mode there was formation of hairline cracks across the fourth cycle [Fig. 5, 7(a)].
joint during the first cycle of loading. Subsequent cy- A major cause for stiffness loss in the specimens is
cles resulted in an increase of the width of the cracks in the slippage of the column bars through the joint and
this region. The specimens' beams were intact at the the pullout of the beam's longitudinal reinforcement
conclusion of the test (Fig. 8). from the joint, especially when the joint yields earlier
during the test and the cracked joint concrete loses its
Effect of inclined reinforcing bars in combination ability to resist the bond forces. Thus the less cracked
with MR andy the beam-column joint concrete is, or the later the
The beam-column connections of Specimens S1 and beam-column joint yields during an earthquake, the less
X 1 , which had very optimal values for design parame- beam and column reinforcement slippage occur.
ters MR and 'Y that involved very low horizontal and By correlating the stiffnesses of the hysteresis loops
vertical joint shear stresses, showed increased strength of specimens S2 and X 2 , expecially near the zero dis-
and ductility and permitted their adjacent beams to de- placement point (pinching of hysteresis loops, Table 2),
velop bending moments 25 and 40 percent higher, re- it can be seen that the inclined bars reduce the slippage
spectively, than their flexural strengths during the first of beam and column longitudinal reinforcement in the
cycle of loading (Table 2). Also, this strength increased joint region by increasing joint strength and ductility
continuously until the fifth cycle in Specimen S1 [Fig. 5, and by reducing the joint concrete cracking.
7(a)] and until the last cycle in Specimen X 1 [Fig. 5, The inclined reinforcing bars helped Specimen X 2
7(b)], while their hysteresis loops show high energy dis- improve its resistance by a change in the failure mode
sipation capacity, minimal decrease in stiffness, and compared with Specimen S2 • It is worth noting that
little pinching. The presence of inclined reinforcing bars Specimen X 2 's joint was intact at the end of the test
in the joint region of Specimen X 1 enhanced its load- while the joint of Specimen S2 had extensive damages
carrying capacity compared with that of Specimen S1 (Fig. 8).
[Fig. 5, 7(a) and (b)]. Besides, the joint of Specimen X 1 Beam-column connections of the other specimens are
was intact at the conclusion of the test, while in Speci- designed with nonoptimal values for parameters MR
men S1 the joint lost part of the concrete cover (Fig. 8). and 'Y. The hysteresis loops of speciments with inclined
As shown in Fig. 5, Specimen X2 maintained its max- bars demonstrated increased strength, ductility, and
imum first cycle load through the eighth cycle of load- energy dissipation capacity and less joint damage com-
ing, while the little strength degradation in the follow- pared with those of their corresponding conventionally
ing cycles is caused by intensive buckling of beam bars reinforced specimens. A representative example of
ACI Structural Journal I January-February 1992 9
exterior beam-column joint's seismic resistance im- cycle, during the last cycle, and the strain gage data of
provement with the presence of inclined bars, while the the vertical shear joint reinforcement for the first cycle,
values of MR and 'Y are nonoptimal, is given by Speci- respectively. It can be seen that the inclined bars yield
men X6 • Fig. 9(a) and (b) show the strain gage data of during the first cy~le but remain active during the whole
Specimens S6 and X6., beam-column joint reinforce- test. The opposite occurs with the vertical joint rein-
ments, respectively, during the first quadrant of load- forcement of Speciment S6 that did not yield. This leads
ing. Fig. lO(a) through (c) show the strain gage data of to the conclusion that inclined bars accept more shear
Specimen X6 's inclined reinforcing bars during the first stresses than vertical joint shear reinforcement, which
remains almost inoperative during the test.
Comparing the beam-column joint hoop reinforce-
ment strain gage data between Specimens S6 and X6 , we
60
find that Specimen's ~·s hoops accept less tension than

r
z
~ 50
specimen 5 6 Specimen S/s hoops, despite the fact that Specimen
a 40 strain gages X6 's beam-column joint was subjected to higher shear
"
.<=
<II 30 location stresses. This was due to inclined bars which relieved
"tJ <D
"
~ 20 Specimen X 6 hoops' tension [Fig. 9(a) and b]. The
0.
c 10 presence of inclined bars enhanced Specimen X/s joint
I strength compared with that of Specimen SJs joint, as
2 0 -1 -2 -3 strain [Ofoo] can be seen from the hysteresis loops (Fig. 6) and the
0 tension strain gage data (Fig. 9). Specimen X6 exhibited less loss
0 compression of strength, less stiffness degradation, less slippage of
(al longitudinal beam and column reinforcement, and
higher energy dissipation capacity than Specimen S6 •
The preceding Specimen ~·s beam-column joint seis-
80 0 (i)®
mic behavior improvement allowed its beam to yield,
70
developing 5 percent more bending moment than its
flexural strength in the third cycle. Specimen S6 's beam
60
z
specimen X 6
did not yield (Table 2). Finally, Specimen X6 showed
~ 50
'-
significantly better mode of failure with less joint dam-
g
.<=
40 strain gages
age compared with that of Specimen S6 (Fig. 8) .
<II location
"tJ 30 <D Comparing the hysteresis loops of Specimens S6 and
fr"c 20 SJ (Fig. 6), it can be seen that the substitution of verti-
cal joint reinforcement with equal percentage of col-
10
I umn moment reinforcement did not significantly im-
2 0 -1 -2 -3 strain [Ofoo1 prove the beam-column capacity, and slippage of beam
0 tension and column reinforcing bars in the joint region. On the
0 compession contrary, Specimen X6 , in which the joint vertical shear
(b)
reinforcement is substituted by inclined bars, demon-
Fig. 9(a) and (b) -Applied shear versus measured strated satisfactory performance.
strain in main beam-column joint reinforcement of Examination of Specimens ~. X7 , and X8 during and
Specimens S6 and X 6 at the conclusion of the test indicated that Specimens

(a) (b) (c)

[I(N] (KN) [KN]

80 ®

-2 - 3stroin[•1•• ] 5 -2 strain[•!..] 2 -2 strain [•!..]

-20

® ®
"'i
@tension
( ) tompression
®
@~
(!)tension
8 compression
-~I
-40

::
-70

-80
®

(!)tension
8 compression

Fig. 10- (a) Strain gage data of Specimen X/s diagonal reinforcement during the
first cycle of loading; (b) during the last cycle of loading; (c) strain gage data of
Specimen S/s vertical shear reinforcement during the first cycle of loading
10 ACI Structural Journal I January-February 1992
X1 and Xs, which had a lower amount of inclined bars 0. sv sx

than specimen x6, had more deterioration of load-car-

rying capacity [Fig. 7(b)] and suffered more damage
than Specimen X6 •

Comparison of the proposed new reinforcing

pattern with the other nonconventional
reinforcing pattern
A comparison of the performance of Specimens P 1, (b) (c)
(a)
Y1, F1, O~o and F2 indicated that significant improve-
ment in the beam-column joint earthquake-resistant Fig. 11 -Mechanisms of shear transfer
mechanical properties was not obtained, despite the use
of higher strength concrete, of stirrups, and of steel fi-
in the following:
ber concrete in the joint. On the contrary, remarkable
1. External beam-column joints with crossed inclined
seismic performance improvement was obtained in
reinforcing bars showed high strength, and no appre-
Specimen F2 , which combined steel fiber concrete with
ciable deterioration after reaching their maximum ca- ·
stirrups in its joint region.
pacity. Also spindle-shaped hysteresis loops were ob-
The joint of Specimen F2 showed more ductility, less
served, with large energy dissipation capacity.
loss of stiffness, and higher energy dissipation capacity
2. Examination of specimens of Types S and X un-
compared to Specimen S6 • The joint of Specimen F2 did
der the influence of the joint shear stresses and flexural
not, however, perform as successfully as the joint of
strength ratio, during and at the conclusion of the tests,
Specimen X6 with the new proposed nonconventional
indicated the following:
pattern of reinforcement (Fig. 6).
Low joint shear stresses in the presence of high flex-
ANALYSIS FOR STRENGTH ural strength ratios resulted in satisfactory perform-
The presence of inclined bars in the joint introduces ance of both exterior beam-column subassemblages,
a new mechanism of shear transfer, in addition to the conventionally reinforced (S) and with inclined bars
two known mechanisms of conventionally reinforced (X).
joints which remain active during the whole test [Fig. Low joint shear stresses in combination with flexural
10(a) and (b)]. Thus, the main mechanisms of shear strength ratio values very close to those recommended
transfer of Specimens X are the following (Fig. 11): by Committee 352 significantly enhanced the behavior
a. The truss mechanism of shear reinforcement; of the subassemblage reinforced with inclined bars. On
b. The mechanism of concrete diagonal compression the contrary, the conventionally reinforced subassem-
strut; blage did not demonstrate satisfactory performance.
c. The truss mechanism of inclined bars. In specimens with high joint shear stresses in combi-
The determination of the portion of shear carried by nation with flexural strength ratio less than the limit of
the third mechanism of the preceding list can be based Committee 352, the performance of conventionally re-
on the statically determinate model whose equilibrium inforced connections did not rate satisfactorily. A sig-
is illustrated by the force polygon [Fig. ll(c)]. This is nificant improvement of the connections reinforced
reasonable for a member subjected to constant shear with inclined bars is observed. The improvement was
and antisymmetric bending, in which the inclined rein- more noticeable as the inclined bar reinforcement in-
forcing bars are in accordance with the shear and mo- creased and as the joint shear stresses were lower.
ment distribution. After first yielding, the shear carried Conventionally reinforced exterior beam-column
by the inclined bars is joints did not succeed in resisting shear stresses higher
than those recommended by Committee 352 (i.e.,
Vsx = 2A, · ./y · sinO (1) 12../T: psi = 1.0../T: MPa). The increase of the flexural
strength ratio did not seem to be able to stop the early
where shear yielding of the connections. Exterior beam-col-
umn jonts with inclined bars resisted horizontal shear
A, = the area of inclined bars
stresses higher than the recommendations of the com-
./y = the yield stress of these bars and
mittee.
8 = the inclination of these reinforcing bars to the col-
3. The presence of inclined bars introduces an addi-
umn axis
tional new mechanism of shear transfer.
More details can be found in Reference 9 where the 4. External beam-column joints with inclined bars
correlation between values predicted by Eq. (1) and also performed considerably better than those with
shear forces measured during the experiments are very conventional reinforcements and nonconventional re-
closed. inforcements.
5. It should be emphasized that in the case of the new
CONCLUSIONS proposed nonconventional way of reinforcement, the
This experimental study c:m the external beam-col- occurrence of diagonal explosive cleavage shear frac-
umn connections reinforced with inclined bars resulted ture was avoided.
ACI Structural Journal I January-February 1992 11
 6. Although the ACI-ASCE Committee 352 design 3. CEB, "Model Code for Seismic Design of Concrete Struc-
gives excellent results, inclined reinforcing bars can be tures," Bulletin d'Information No. 165, Paris, Apr. 1985.
4. ACI-ASCE Committee 352, "Recommendations for Design of
used to improve seismic resistance in beam-column Beam-Column Joints in Monolithic Reinforced Concrete Structures
joints that do not conform to Committee recommen- (ACI 352R-85)," Amerit:an Concrete Institute, Detroit, 1985, 19 pp.
dations. 5. Paulay, T.; Park, R.; and Phillips, M. H., "Horizontal Con-
struction Joints in Cast-in-Place Reinforced Concrete," Shear in Re-
inforced Concrete, SP-42, American Concrete Institute, Detroit 1974,
ACKNOWLEDGMENTS pp. 599-616.
The work reported herein is based on A. G. Tsonos' PhD thesis 6. Minami, K., and Wakabayashi, M., "Seismic Resistance of Di-
submitted to the Civil Engineering Department of Aristotle Univer- agonally Reinforced Concrete Columns," Proceedings, Seventh
sity, Thessaloniki, Greece, and prepared under the supervision of World Conference on Earthquake Engineering, Istanbul, 1980, Ke-
Professor G. Gr. Penelis and Assistant Professor I. A. Tegos. This laynak Printing Co., Ankara, V. 6, pp. 215-222.
research program was sponsored by Aristotle University and the sup- 7. Tegos, 1., and Penelis, G. G., "Seismic Resistance of Short
port is gratefully acknowledged. Columns and Coupling Beams Reinforced with Inclined Bars," ACI
JOURNAL, Proceedings V. 85, No 1, Jan.-Feb. 1988, pp. 82-88.
8. Tegos, I. A.; Tsonos, A. G.; and Penelis, G. Gr., "Seismic Re-
NOTATION sistance of Diagonally Reinforced Concrete Beam-Column Joints,"
M. sum of the flexural capacity of columns to that of beam Ninth World Conference on Earthquake Engineering, Tokyo-Kyoto,
V joint shear stress as a multiple of .JJ: Japan 1988, 6 pp.
a stress multiplier for flexural reinforcement = 1.25 for earth- 9. Tsonos, A. G., "Contribution to the Seismic .Behaviour's Im-
quake loading provement of Exterior B/C Joints of Reinforced Concrete Frame
a hysteresis loop pinching Structures," PhD thesis, Aristotle University of Thessaloniki, 1990,
p, transverse reinforcement ratio, percent Appendix 46, V. 11, 405 pp. (in Greek)
ld. development length of hooked bars measured from the face 10. ACI Committee 544, "State-of-the-Art Report on Fiber Rein-
of the column core to back side of the hook forced Concrete (ACI 544.1R-8Z)," American Concrete Institute,
/d development length for deformed bars in tension Detroit, 1982, 22 pp.
<P bar diameter 11. Henager, C. H., "Steel Fibrous, Ductile Concrete Joint for
f: compressive strength of concrete Seismic Resistant Structures," Reinforced Concrete Structures in
h. total depth of beam Seismic Zones, SP-53, American Concrete Institute, Detroit, 1977,
b. width of beam pp. 371-386.
h, total depth or width of square column 12. Tsonos, A. G.; Tegos, I. A.; and Penelis, G. Gr., "Seismic
A, area of inclined bars Behaviour of Exterior Beam-Column Joints of Conventional or Fi-
J: yield stress of inclined bars bre-Reinforced Concrete," COMP '88 Phase Interaction in Compos-
fl inclination of inclined bars to the column axis ite Materials, Second International Symposium, Patra, Greece, Aug.
1988, 6 pp.
13. Ehsani, M. R., and Wight, J. K., "Exterior Reinforced Con-
CONVERSION FACTOR crete Beam-to-Column Connections Subjected to Earthquake-Type
1 psi = 0.0069 MPa Loading," ACI JouRNAL, Proceedings V. 82, No 3, May-June 1985,
pp. 343-349.
14. Ehsani, M. R., and Wight, J. K., "Effect of Transverse Beams
REFERENCES and Slab on Behavior of Reinforced Concrete Beam-to-Column
l. ACI Committee 318, "Building Code Requirements for Rein- Connections," ACI JOURNAL, Proceedings V. 82, No 2, Mar.-Apr.
forced Concrete (ACI 318-83)," American Concrete Institute, De- 1985, pp. 188-195.
troit, 1983, Ill pp. 15. Scarpas, A., "Inelastic Behavior of Earthquake Resistant Re-
2. NZS 3101:1982, "Code of Practice for the Design of Concrete inforced Concrete Exterior Beam-Column Joints," Report No. 81-2,
Structures," Standards Association of New Zealand, Wellington, Department of Civil Engineering, University of Canterbury, Christ-
Part 1, 127 pp., Part 2, 156 pp. church, Feb. 1981, 84 pp.

12 ACI Structural Journal I January-February 1992