•

IBI
PRECAST/PRESTRESSED
CONCRETE INSTITUTE
Daniel P. Jenny Research Fellowship

Moment Resistant Connections

in Precast Concrete Frames
for Seismic Regions
Jay E. Ochs, P.E.
Project Engineer
Sundt Corporation
Tucson, Arizona
The concept of relocating the plastic hinge region of precast
concrete frames for construction in seismic regions is
examined. Five beam-to-column subassemblies were tested
under simulated earthquake-type loading. The connections
included steel plates or angles embedded in the columns and
beams which would facilitate field erection. The column axial
load was held constant during the test while the free end of
the beam was subjected to reversed cyclic displacements.
The failure modes, strength and ductility of the test specimens
using various connection details are compared. It is con-
cluded that properly designed connections with relocated
plastic hinges can perform in a satisfactory manner. The
reduced welding requirements for these details result in
improved economy of construction.

he design of ductile moment re- ucts are of high quality and can be

T sisting beam-to-column connec-

Mohammad R. Ehsani, Ph.D.
Associate Professor of Civil
Engineering and Engineering
Mechanics
University of Arizona
Tucson, Arizona
tions for monolithic reinforced
concrete structures has been studied
by many researchers over the past
three decades. As a result of these
studies, design recommendations for
such conn~ctions have been developed
by ACI-ASCE Committee 352.'
One disadvantage of monolithic
beam-to-column connections is that,
for any building more than one story
high, elaborate formwork and supports
are necessary for construction. Precast
concrete members, on the other hand,
can be fabricated in a plant under con-
trolled factory conditions. The prod-
erected quickly. Consequently, precast
concrete frames are an economical al-
ternative to cast-in-place construc-
tions, especially when dead load ,con-
stitutes the dominant loading of the
structure and simple connections are
used.
For frames subjected to lateral load-
ing , beam-to-column conriections
must be capable of resisting large
forces and displacements. Due to the
difficulties in providing such connec-
tion details, the development of eco-
nomical moment resisting connections
for precast concrete frames has been
identified as one promising area which

64 PCI JOURNAL
requires further investigation. 2 The
large amount of research and number
of publications that have appeared in
recent years clearly indicate the inter-
est of the industry and the engineering
profession in addressing some relevant
questions.3- 17
In accordance with strong-column
weak-beam philosophy, most build-
ings are designed so that flexural
hinges are formed at the ends of the
beams, near the column faces. This
has been the primary objective of most
research conducted on beam-to-
column connections for precast con-
crete frames, including the work of
Pillai and Kirk; and Bhatt and Kirk. 7
These studies were conducted to de-
velop a ductile connection detail.
Although the behavior of the tested
connections was satisfactory, the con-
struction of these specimens requires
significant welding of the beam and
column reinforcement. The cost and
quality control associated with exces-
sive welding diminishes some of the
inherent advantages of precast concrete
construction. For the ideal connection,
therefore, welding - especially field
welding - must be minimized.
The concept of relocating the beam
plastic hinges away from the face of
the column is relatively new. The ad-
vantages of such details are the reduc-
tion of strength and ductility demands
on the connection. In monolithic struc-
tures, this will result in a reduction of
confining reinforcement in the con-
nection region.
The first investigations of these con-
nection details were carried out in New
Zealand. The results of these studies
eventually led to the inclusion of
guidelines for the relocated plastic
hinge concept in the New Zealand
Code of Practice for the Design of
Concrete Structures. 18 Recent studies
by Abdel-Fattah and Wight, 19 and Al-
Haddad and Wight 20 indicated their
success in relocating the plastic hing-
ing zone away from the face of the col-
umn and into the beams for monolithic
frames.
The studies by Frencb et al. 12 were
aimed at finding precast concrete con-
nections which would respond favor-
ably to load reversals. The research
provided promising connection de-
tails, including a threaded reinforcing
September-October 1993
Table 1. Description of the test specimens.
Specimen description
Specimen Type of Hinge
designation construction location
MR Monolithic Relocated
Pl Precast Column face
PRJ Precast Relocated
P2 Precast Column face
PR2 Precast Relocated
Note: I psi = 6.895 MPa.
bar connection and a composite sec-
tion comprised of precast and mono-
lithic connections. Recent studies have
evaluated connections and new sys-
tems which can be used in regions of
high seismicity, 1.. 16 including also the
use of debonded tendons to provide
the connection between the beams and
columns. 17
RESEARCH OBJECTIVE
Because earlier tests 3- 17 had indicated
that ductile precast beam-to-column
connections were cost prohibitive,
an alternative approach would be to
design precast concrete connections
with relocated plastic hinges. This
would result in a reduced demand for
strength and ductility in the connec-
tion region. Therefore, the two objec-
tives of this study were to develop
economical ductile moment resistant
connections for precast concrete mem-
bers and to study the effects of locat-
ing the plastic hinge away from the
beam-to-column connection.
DESCRIPTION OF
TEST SPECIMENS
Five specimens, as described in
Table 1, were tested for this study.
The specimens were designed in ac-
cordance with the provisions of the
ACI Building Code21 and the latest de-
sign recommendations of ACI-ASCE
Committee 352. 1 This report contains
recommendations for joint shear
stresses, transverse reinforcement, and
development length of beam and col-
umn longitudinal reinforcement within
the joint. These guidelines are for
Beam longitudinal
reinforcement
Concrete
At column Strength
face At 1.5 d (psi)
4#7, 2#6 2#7 4630
4#7 4#7 4700
4#7, 2#6 2#7 4930
4#6 4#6 4780
4#6, 2#5 2#6 4780
monolithic connections . However,
since no specific design recommenda-
tions exist for ductile precast connec-
tions, the recommendations for mono-
lithic frames were used in this study
when applicable.
Each specimen is designated by one
or two letters followed by a number as
listed in Table 1. The first letter indi-
cates the type of construction, M for
monolithic and P for precast. The let-
ter R denotes that the plastic hinge
was relocated from the face of the col-
umn into the beam. The numerals 1
and 2 refer to the two different details
used for the precast connections.
The test column had a cross section
of 16 x 16 in. (406 x 406 mm) and a
height of 10ft (3.05 m). Column rein-
forcement consisted of eight No. 8 or
eight No. 7 Grade 60 longitudinal bars
and No. 4 Grade 60 closed ties spaced
at 3 in. (76 mm) within the joint and
spaced at 4 in. (102 mm) elsewhere.
The beams were 14 x 24 in. (356 x
610 mm) in cross section and ex -
tended 52 in. (1.32 m) from the col-
umn face. No. 4 Grade 60 closed stir-
rups were provided on 4 in. (102 mm)
centers for all beams.
Flexural reinforcement for each
beam is listed in Table 1. All beams
had equal compression and tension re-
inforcement. Throughout this paper,
reference is made to the top and bot-
tom of the beams . These locations
were selected as shown in Fig. 1.
The specimens were designed using
the strong-column weak-beam con-
cept. The plastic hinges were expected
to develop at the end of the beam near
the column for Specimens Pl and P2
and at a distance of one beam depth,
65

A
Ill (_L
- I
~ d
~
1.5d
Mn=1.25 Ma
Ma
[ I
NOT TO SCALE
Fig. 1. Schematic of beam reinforcement detail for specimens with relocated
plastic hinges.
i.e., 21 in. (533 mm) away from the
face of the column in Specimens MR,
PRl and PR2.
The specimens with the relocated
plastic hinges were designed taking
into account several new concepts in-
cluding recommendations from Abdel-
Fattah and Wight. 19 Two beam cross
sections were designed, one at the col-
umn face and the other at one effective
beam depth away from the column
face. It was necessary to design these
two sections so that when beam hing-
ing occurred at the intended location,
no excessive yielding would occur at
the column face.
The beam section at the column face
was designed to have a nominal flexu-
ral capacity of about 25 percent larger
than the maximum anticipated acting
moment. Fig. 1 helps explain this con-
cept. The tension and compression
bars and the intermediate longitudinal
reinforcement were used in designing
and calculating the capacity of this
section.
For all specimens, equal amounts of
66
(EXPECTEDRE LOCATED
HINGINGZON E
/OPO FBEAM
Td
_j_
I \BOTTOM OF BEAM
. The beam tension reinforcement con-
Mn=Ma
J the top and bottom beam longitudinal
tension and compression steel were
used for the beams. The intermediate
reinforcement was extended a distance
of one-and-a-half times the beam
depth from the column face. The inter-
mediate reinforcement area was about
35 percent of the tension or compres-
sion reinforcement.
The ratio of the nominal moment
capacity of the beam section at the de-
sired plastic hinging location to that at
the end of the beam should be selected
based on the expected moment dia-
gram for the beam. Here, an attempt
was made to keep the ratio of the
beam capacity at the hinge to the beam
capacity at the column face less than
or equal to 0.55. A similar ratio has
been used and recommended by other
researchers. '9 ·20 To achieve this ratio, it
was necessary to terminate some of
the beam tension and compression
steel in the specimens with relocated
plastic hinges.
The nominal moment capacity of
the beam section at the relocated hing-
ing zone was calculated ignoring the
intermediate bars. In other words, it
was assumed that the extension of the
reinforcement for a distance of half
the beam depth is not long enough to
develop the full tensile strength of
these bars.
Specimen MR
In an attempt to draw comparisons be-
tween precast and monolithic frames, a
monolithic specimen was constructed
and tested as a control specimen (MR).
sisted of four No. 7 bars placed at 3 in.
(76 mm) from the extreme fiber. In ad-
dition, four No . 6 intermediate bars
were symmetrically placed at 6 in .
(152 mm) from the mid-height of the
beam. In an effort to relocate the plas-
tic hinge, all intermediate and two of
bars were terminated at a distance of
one-and-a-half times the beam depth
from the column face.
Specimens P1 and PR1
The connection detail for the top
and bottom of the beam was identical
for these two specimens. As shown in
Fig. 2, the connection for these two
specimens consisted of two fabricated
steel T-sections embedded in the col-
umn. Each T -section had three holes
to allow for the passing of column
longitudinal bars. Four No. 7 standard
90 degree hooks were welded to the
T -sections to provide adequate an-
chorage of the plate within the joint.
The beam end included two large
steel angles to which the longitudinal
reinforcement was welded. The welds
were % in. (10 mm) flare-bevel welds,
3 in. (76 mm) long on both sides of
the reinforcement. The steel angles
and T-sections are shown in Fig. 3.
The intermediate reinforcement for
Specimen PRl was provided in the
form of U-shaped No. 6 bars. Each bar
was placed in a horizontal plane and
its base welded to the steel angles with
% in. (10 mm) flare-bevel welds over
the full length of the base. After the
beam and columns were cast and
cured, the beam angles were welded to
the column T-sections with ~ in. (13
mm) fillet welds over the full width of
the beam to complete the connection.
PCI JOURNAL
 21 "
24"

#4CLOSED
STIRRUPS AT
4"0.C.

#6 U-SHAPED BARS
- TOP AND BOTTOM
I
#7BARSCUTT02AT 16"
( 1.5d- TOP AND BOTTOM ( TOP AND
3 3/8 I BOTTOM
I.. 1.sd
31.5 11
#4 HOOPS AT 3" O.C.
WITHIN JOINT
• ·

NOTTOSCALE
-~~_13"
16" ~
-~ _______.
.__

Fig. 2. Reinforcement detail for Specimen PR1. Note: 1 in. = 25.4 mm.

Specimens P2 and PR2
The connection detail for Specimens
P2 and PR2 was developed in collabo-
ration with the engineers of the Tanner
Companies and Stanley Structures,
both in Phoenix, Arizona. The detail
was designed to provide adequate
strength and to facilitate actual con-
struction in the field . Details of the
connection are shown in Fig. 4.
For the column, a T-section similar
to that of Specimens P1 and PR1 was
used on the top. However, for the
lower side, a straight plate which ex-
tended from the column face was uti-
lized. The plate would serve as a beam
seat in the field. In addition, the size
of the grouted shear key could vary
slightly to accommodate field toler-
ances such as column misalignment or
slight variations in beam length.
Each of the top bars in the beam
was welded to a steel angle with Ys in.
(10 mm) flare-bevel welds, 4 in. (102
5"X 5" X 3/8"
I1"LOOO
T- SECTION ANGLE
Fig. 3. Detail of steel angle and
T-section used in Specimens P1
and PR1. Note: 1 in. = 25.4 mm.
mm) long on both sides of the rein-
forcing bar. The ends of these angles
were welded to the vertical face of thy
steel T -sections in the columns with
Y. in . (6 mm) fillet welds all around
and the surrounding area was grouted.
On the bottom side of the beam, the
bars were welded to a rectangular
steel plate. The plate was later welded
to the column with a 716 in. (8 mm) fil-
let weld.
Due to a lack of availability of
larger sized weldable reinforcing bars,
the design of the beams for Specimens
P2 and PR2 incorporated No. 6 and
No. 5 bars for the main reinforcement
(Table 1). The intermediate reinforce-
ment for Specimen PR2 was provided
in the form of U-shaped bars placed in
a vertical plane so that when one leg
of the bars was in tension, the other
would be in compression.
TEST SETUP AND
INSTRUMENTATION
The specimens were tested in a steel
reaction frame as shown in Fig. 5. The
column portion of the specimens was
placed horizontally in the frame. The
specimens were held in position by
means of steel clevises which were
bolted to anchor bolts cast at the ends
of the beam and column. The resulting

September-October 1993 67
 12.5/8" X6"XI6"
NOT TO SCALE
Fig. 4. Connection details for Specimens P2 and PR2. Note: 1 in. = 25.4 mm.
Fig. 5. Test setup.
pinned connections were intended to
represent the inflection points of the
frame. Using a hydraulic jack, an axial
load of 20 kips (89 kN) was applied to
the column. The axial load was kept
constant throughout each test.
68
4#6 TOP
AND BOTTOM
# 5u-SHAPED
BARS
Cyclic shear was then applied to the
free end of the beam. The loading his-
tory selected for the tests consisted of
one cycle with a maximum displace-
ment of 0.5 in. (13 mm), two cycles
at 1 in. (25 mm), two cycles at 2 in.
8 I
·-
10.25"
·-
10.25"
7 2
0}•"
1. 23.0" ~I
Fig. 6. Location of displacement
transducers in specimens.
Note: 1 in. = 25.4 mm.
(51 mm), two cycles at 3 in. (76 mm)
and two cycles at 4 in. (102 mm). The
initial displacement cycle of 0.5 in.
(13 mm) corresponds to small drifts in
the elastic range, while the repeated
cycles at larger displacements were
intended to evaluate the ability of the
specimens to maintain their load car-
rying capacity and stiffness under
large inelastic deformations. This
loading history was followed for Spec-
imens MR, Pl and PRl. However,
because of extensive cracking around
the beam plates in Specimens Pl and
PRl, the loading cycles were modi-
fied to apply displacements in ~ in.
(13 mm) increments for Specimens P2
and PR2.
Twelve to 14 electrical resistance
strain gauges were placed on beam
and column longitudinal and trans-
verse reinforcement. Moreover, in
order to monitor the joint deformation
and beam rotation during the loading
sequence, eight displacement trans-
ducers were placed on the specimens
as shown in Fig. 6. Throughout the
tests, loading was temporarily stopped
while the load, displacement transduc-
ers and strain gauges were automati-
cally read and recorded using a high
speed data acquisition system.
MATERIAL PROPERTIES
The concrete for this experimental
program, obtained from a ready-mix
plant, had a 28-day design compres-
sive strength of 4000 psi (28 MPa).
The mix proportions per cubic yard
of co ncrete are as fo llows: aggre-
PCI JOURNAL
Fig. 7. Specimen MR after two 2 in. (51 mm) displacement Fig. 9. Specimen P1 after two 2 in. (51 mm) displacement
cycles. cycles .

.eo -40 -20 +20 +40 + 60 (mml .eo -60 -40 -20 + 20 + 40 +60 +80(mmJ
z~ i
0
0
N
8N
+ +

~ ~
+ +

0 0

~ ~
8 8
"'! "'!

- 79 DISPLACEMENT CINJ - 79 DISPLACEMENT CINI

Fig. 8. Load vs. displacement for Specimen MR. Fig. 10. Load vs. displacement for Specimen P1.
Note: 1 kip= 4.45 kN; 1 in. = 25.4 mm. Note: 1 kip= 4.45 kN ; 1 in. = 25.4 mm.

gate with a maximum size of :Y. in.
(19 mm) , 1780 lbs (7.92 kN); sand,
1220 lbs (5.43 kN); Type I cement,
500 lbs (2.22 kN) ; fly ash, 115 lbs
(0.51 kN) ; and water 310 lb s (1.38
kN). The actual concrete compressive
strength for each specimen at the time
of the test is given in Table 1.
The reinforcement was Grade 60
weldable steel. The following average
yield stresses were obtained from
coupon tests of the reinforcement:
No . 4 bars , 77.5 ksi (534 MPa) ;
No. 5 bars, 72.3 ksi (498 MPa); No. 6
bars, 69.0 ksi (476 MPa) ; No. 7 bars,
65.7 ksi (453 MPa); and No. 8 bars,
62.7 ksi (432 MPa).
All plates and angle s used were
Grade A36 steel. The column plates
for the four precast concrete speci-
mens and the beam plates for Speci-
mens P2 and PR2 were % in. (16 mm)
thick. The large steel beam angles
used in Specimens P1 and PR1 were
5 X Ys X 14 in. (127 X 10 X 356 mm)
long. All welding was done using
E7018 electrodes.
TEST RESULTS
Specimen MR
Thi s specimen wa s constructed
monolithically and was designed to
have the plastic hinge develop away
from the column face. At 1 in. (25
rnm) displacement, a flexural crack of
about 0.2 in. (5 mm) in width formed
on the tension side of the beam. Tbis
crack was located at about 30 in. (760
mm) from the column face. It also
indicated that the pl astic hinge was
beginning to develop at about one-
and-a-half times the beam depth from
the column face. At a displacement of
-2 in. (-51 mm) , additional tension
cracks formed on the tension face of
the beam. The largest of these cracks
was 0.25 in. (6 mrn) wide. In addition,
the ten sio n cracks in the column
within the joint region were becoming
more noticeable.
After two 2 in. (5 1 mrn) cycles, the
specimen was severely cracked, as can
be seen in Fig. 7. In the following two
3 in. (76 mrn) cycles, further cracking
of the beams and columns was ob-
served. Load vs. displacement for
Specimen MR is plotted in Fig. 8. This
plot does not show the second 3 in .
(76 mrn) cycle because the test was in-
terrupted due to equipment problems
and the hysteresis response of the last
cycle was not plotted.
Specimen P1
Specimen Pl was constructed using
precast concrete members with the

September-October 1993 69
Fig. 11. Specimen PR1 after two 3 in. (76 mm) displacement Fig. 13. Beam plate in Specimen P2 at conclusion
cycles. of test.

r---~--~--~---.--~--~--~----~
.eo -40 ·20 +20 + 40 + 80 {mm)
~
.eo -40 -20 + 20 + 40 + 80 {mm)
8 ..
.
0
0
+
41il +

~
+
0

~
8
"!

-78 DI SPLACEMENT ( INl -Siil DISPLACEMENT tiN>

Fig. 12. Load vs. displacement for Specimen PR1. Fig. 14. Load vs. displacement for Specimen P2.
Note: 1 kip= 4.45 kN; 1 in. = 25.4 mm. Note: 1 kip= 4.45 kN; 1 in. = 25.4 mm.

beam plastic hinge expected to form at
the column face. Several flexural
cracks were observed after the first
1 in. (25 mrn) cycle. At the end of the
second 2 in. (51 mrn) cycle, the crack-
ing was very extensive (as shown in
Fig. 9) and the column plates were be-
ginning to separate from the concrete.
During the first 4 in. (102 mrn) cycle,
a loud popping sou nd was heard as
one of the beam bars broke and the
test was terminated shortly following
fracture of the bar. This bar broke pre-
maturely, probably due to the brittle-
ness introduced by the welding of the
bar to the plate. A plot of load vs. dis-
placement for Specimen Pl is shown
in Fig. 10.
Specimen PR1
Specimen PR1 , fabricated using pre-
cast concrete components, was de-
signed to relocate the plastic hinge
away from the face of the column.
After the first 1 in. (25 mm) cycle,
there were several small cracks with-
in the beam-to-column joint. As the
test proceeded through the two 2 in.
(5 1 mm) cycles , cracking was ob-
served throughout the beam and col-
umn. There were also noticeable cracks
forming around the column plate and
beam angles. The most severe crack-
ing was at the plastic hinge regio n,
near 1.5d, where the cracks were up to
0.20 in. (5 mrn) wide. The cracks
within the joint, for example, were
only about 0.01 in. (0.25 mm) wide.
During the first 3 in. (76 mrn) cycle,
a large tensile crack formed near the
steel angle on the beam. After the sec-
ond 3 in . (76 mm) cycle, the plastic
hinge was well.defined as can be seen
in Fig. 11, and the crack near the beam
angle was over 0.25 in. (6 mrn) wide.
During the first 4 in . (102 mm) cycle,
at a displacement of 3 ~ in. (89 mm), a
loud popping sound was heard. This
was due to the breaking of one of the
reinforcing bars in the beam near the
face of the column, and the test was
ended. A plot of load vs. displacement
for Specimen PR1 is shown in Fig. 12.
Specimen P2
As explained earlier, the final two
precast concrete specimens incorpo-
rated a different connection detail in
addition to longitudinal No. 6 beam
bars, as compared to the No . 7 bars
used in the previous three specimens.
Specimen P2 was de signed without
any intermediate reinforcement.
During the first 1 in. (25 mm) dis-
placement cycle, two cracks devel-
oped, both emanating from the top of
the angle embedded in the beam. One

70 PCI JOURNAL

Fig. 15. Da mage to a rea arou nd plate in Specimen PR2 at conclus ion of test.
crack extended from the top of the
angle at approximately 45 degrees to
the column, while the other moved
from the top of the angle into the
beam along the longitudinal beam
bars. These cracks continued to widen
through subsequent cycles until the
eventual failure of one of the welds
along the side of the plate.
Following two cycles at 1~ in. (38
mm), a beam bar welded to the plate
side of the beam fractured at a dis-
placement of 1.58 in. (40 mm). Fig. 13
shows the extensive cracking and loss
of concrete around the beam plate. In
an attempt to obtain as much addi-
Septem ber-October 1993
tiona! data as possible, the specimen
was pushed in the other direction to a
displacement of 1.85 in. (47 mm)-at
which time a weld failure occurred be-
tween the beam angles and the column
T-section. The plot of load vs. dis-
placement for Specimen P2 is given in
Fig. 14.
Specimen PR2
Specimen PR2 was identical to
Specimen P2 with the exception that
intermediate reinforcement was pro-
vided and some of the main reinforce-
ment was terminated at a distance of
l.Od in an attempt to relocate the plas-
tic hinge region in the beam.
During the first 1 in. (25 mm) dis-
placement cycle, two cracks devel-
oped at the beam plate almost identi-
cal to the previous specimen. In
addition, substantial cracking occurred
at approximately 20 in. (508 mm)
from the base of the beam in an X-
shaped pattern, which indicated the
formation of a plastic hinge at this lo-
cation in the beam.
Following two cycles at I ~ in. (38
mm), an attempt was made to displace
the specimen to 2 in. (51 mm), but at a
displacement of 1.79 in. (45 mm) one
of the beam bars on the plate side of
the beam fractured . It was decided to
conclude the test at this point due to
the violent nature of the failufe. Fig.
15 shows extensive cracking around
the beam plate following the test. The
load vs. displacement plot for Speci-
men PR2 is shown in Fig. 16.
COMPARISON OF
TEST RESULTS
Moment Capacity
One measure of performance of a
test specimen is to compare its mea-
sured flexural capacity with the nomi-
nal moment capacity based on the ac-
tual measured material properties. The
test is considered successful, at least
as far as overall moment capacity is
concerned, if the specimen is able to
resist this design load. The ratio of the
maximum measured moment to the
nominal moment capacity for Speci-
mens MR, Pl , PRl, P2 and PR2 are
1.31 , 0.98, 1.04, 1.07 and 1.20, respec-
tively. All tested specimens, except
Specimen PI, resisted moments which
were larger than their nominal mo-
ment capacities.
Energy Dissipation
A measure of the energy dissipated
by a specimen is represented by the
area enclosed within the load vs. dis-
placement curves for any given load
cycle. The energy dissipation is a mea-
sure of the ductility of the specimen.
The energy dissipated at each cycle is
normalized with respect to the product
PY Lly. This normalization is needed to
71

.---~~--~--r-_.--~~----.i
-60 -40 ·20
-50
Fig. 16. Load vs. displacement for Specimen PR2.
Note: 1 kip= 4.45 kN; 1 in. = 25.4 mm.
Table 2. Energy dissipated per cycle .
Cycle
Number MR
I -
2 0.27
3 0.18
4 1.21
5 0.89
6 1.62
7 2.02
Total 6.19
eliminate the effect of the members
having different amounts of flexural
reinforcement. The normalized amount
of energy dissipated for each load
cycle is shown in Table 2.
Table 2 shows that the energy dissi-
pated by each of the two precast Spec-
imens P1 and PR1 was higher than
that of the monolithic Specimen MR.
Although Specimen MR resisted more
load, Specimens P1 and PRl were
able to dissipate more energy. Speci-
mens P2 and PR2 did not dissipate as
much energy as Specimen MR. This is
because of the excessive cracking and
rapid deterioration of the concrete sur-
rounding the beam seat plate.
Strain Gauge Data
As mentioned earlier, strain gauges
were attached to the steel reinforcement
72
0 25A 50.8 78.21mml
+ 20 + 40 + 80 lmml
~
~~------~------~--------~--~
---MR
+ P1
---PR1
P2
DISPLACEMENT <IN>
0 0.5 1 1.5 2 2.5 3 3.5
MAXIMUM DISPLACEMENT (IN.)
Fig. 17. Maximum strain in beam bars at different cycles of
loading. Note: 1 in. = 25.4 mm.
were located near the column face
Specimen
where the plastic hinge developed.
Specimens P2 and PR2 did not show
PRl PI P2 PR2 the same response because the tests of
these specimens were stopped before
0.04 0.05 0.05 0.10
the 3 in. (76 mm) cycles, which is
0.31 0.24 0.16 0.39 when extensive cracking usually oc-
curred in the beams.
0.2 1 0.14 0.47 0.26
The plots of the strain in the inter-
1.41 1.00 0.43 1.01 mediate beam bars are shown in Fig.
19. Comparing Specimens MR and
1.05 0.68 1.14 0.78 PR1, it was concluded that the slightly
3.32 2.34 1.32 0.85
higher strains in Specimen MR were
due to the continuity across the beam-
2.81 1.95 0.8 1 - to-column interface in that specimen.
After reviewing the data from Speci-
9.15 6.40 4.38 3.39
men PR2, it was shown that by contin-
uing the intermediate reinforcing bars
at locations of interest. The strains in from the tension side into the com-
the main beam bars are shown in Fig. pression side of the beam, the interme-
17. The gauge for Specimen PR2 did diate bars were able to develop more
not work and, thus, there is no plot for resistance than was expected from the
this specimen. The plots are very simi- results of the other two tests. This
lar and , as expected, the strains in means that the connection detail used
Specimens MR and PR1 were less than in Specimen PR2 (see Fig. 4) proved
those of Specimen Pl. This is due to to be a better way to anchor the inter-
the presence of intermediate reinforce- mediate beam bars.
ment in Specimens MR and PRl. The plots of the strain in the column
The strains in the beam stirrups bars are shown in Fig. 20. Although
closest to the column are shown in the plots were similar, the monolithic
Fig. 18. It was interesting to note that specimen, MR, resisted more strain
the strains in Specimen MR were only than the precast concrete specimens.
a fraction of the strain in all the pre- This was consistent with the load vs.
cast concrete specimens. Specimen P1 displacement plots which showed that
had much higher strains than other for the same amount of deflection, the
specimens. This can be expected be- monolithic specimen (MR) resisted
cause Specimen P1 did not have the slightly more load than the precast
plastic hinge relocated away from the concrete specimens (Specimens PRl
column face and the stirrups, which and P1). For Specimens P2 and PR2,
had strain gauges attached to them, the damage was concentrated in the
PCI JOURNAL
 0 25.4 50.8 78.21mml 0 25.4 60.8 78.21mml
1800
-+- -+-
MR

~
MR
~
~
P1 P1

!
---
PR1
-+-
z ---
PR1
-+-
P2
E _,.... E P2
_,....
~ PR2 ~ PR2
~ u
:I i

1 1.5 2 2.5
MAXIMUM DISPLACEMENT (IN.)
3.5 0.5 1 1.5
db2 2.5 3 3.5
MAXIMUM DISPLACEMENT (IN.)

Fig. 18. Maximum strain in beam stirrups at different cycles Fig. 20. Maximum strain in column bars at different cycles of
of loading. Note: 1 in. = 25.4 mm. loading. Note: 1 in. = 25.4 mm.

0 26.4 60.8 78.2 lmml 0 26.4 60.8 78.21mml

1800
~'T---------T---------~--------~--~ -- -+-
1800 MR

1400 P1

z
1400
z
---
PR1

~ ~
1200
1000 P2
II)
1000
0
a: ~ 800 PR2
u 800 u
i i 800
800
400

0.5 1 1.5 2 2.5 3 3.5 0.5 1 1.5 2 2.5 3 3.5

MAXIMUM DISPLACEMENT (IN.) MAXIMUM DISPLACEMENT (IN.)

Fig. 19. Maximum strain in intermediate beam bars at Fig. 21 . Maximum strain in column hoops at different cycles
different cycles of loading. Note: 1 in. = 25.4 mm. of loading. Note: 1 in. = 25.4 mm.

beams and, consequently, the strains
in the column bars remained well
within the elastic range.
The plots of the strain in the column
hoops are shown in Fig. 21. The
strains were similar for all specimens.
Specimen Pl consistently had slightly
higher strains, but that is because it re-
sisted more load than the relocated
Table 3. Joint shear deformation.
Displacement
(in.) MR
I 3.16
2 5.14
3 7.60
* At maximum displacement of 1.66 in. (42 mm).
plastic hinge Specimens MR and PRl. Displacement Transducer Data
For the same reason, Specimen P2 had
more strain on the column hoops than For each specimen, eight displace-
Specimen PR2 as the test proceeded ment transducers were mounted on
past the second 2 in. (51 mm) cycle. the beam-to-column connection and
The data clearly indicate that joint re- on the sides of the beam as shown in
inforcement similar to that suggested Fig. 6. The data from these sensors
for monolithic construction 1 is ade- were used to evaluate the joint shear
quate for precast connections. deformations for each specimen. The
total joint shear deformation is de-
fined as the sum of the average verti-
cal and average horizontal compo-
nents. Fig. 22 shows the deformed
Joint shear deformation (10·' radians)
joint configuration. Table 3 gives the
PRl Pl P2 PR2 total joint shear deformations at
recorded displacements. The higher
1.80 0.95 2.04 1.40 the total joint shear deformation, the
4.85 2.95 3.76 2. 18*
more cracks are formed in the beam-
to-column connection.
5.99 5.35 - - It can be observed that the concen-
tration of damage in the joints of the

September-October 1993 73
Fig. 22. Deformed joint configuration .
precast concrete members (Specimens
PRl and Pl) was substantially less (21
and 30 percent, respectively) than the
similarly constructed monolithic spec-
imen (MR). In other words, most of
the deformation in the precast concrete
frames, especially at lower load levels,
was concentrated in the beams away
from the columns, resulting in smaller
joint deformations.
Crack Patterns and
Failure Modes
Throughout the tests, crack develop-
ment was observed, marked and pho-
tographed at regular intervals and at
points of interest to provide an accu-
rate record of crack development. The
crack patterns are shown in several
figures. These patterns show that the
details used were effective in remov-
ing the plastic hinges from the faces of
the columns for Specimens MR, PRl
and PR2. The cracking within the
beam-to-column connections and
within the columns for each specimen
appear similar, but the displacement
transducers showed the differences
previously discussed.
The eventual failure of each precast
concrete specimen was caused by the
fracture of one of the beam bars. This
indicates a localized problem near the
beam plates or angle. Fig. 23 shows
this crack for Specimen PRl; Fig. 13
shows this crack for Specimen P2. For
both details, the steel plate or angle ro-
tated excessively. This, in turn, is be-
lieved to have induced additional
stresses to the beam bars, which when
added to the tensile stresses from the
74
Fig. 23. Specimen PR1 at conclusion of test.
beam bending action, could have con-
tributed to the specimens failing ear-
lier than desired. In addition, the beam
bars may have become brittle due
to their welding to the steel angles
or plates, which may have contribu-
ted to the undesirable failure modes
experienced.
CONCLUSIONS
The two areas studied in this experi-
mental program were the relocation of
plastic hinging zones and the ductility
of moment-resisting precast concrete
connections_. The research objectives
were to establish plastic hinges away
from the column face and to develop
moment resistant connections which
would perform similarly to monolithi-
cally cast connections. The overall
performance of the test specimens was
found to be satisfactory. The follow-
ing conclusions are drawn from the re-
sults of this investigation:
1. The beam plastic hinging zone
can be relocated away from the col-
umn face by using intermediate layers
of longitudinal reinforcement and
some tension and compression bars at
a predetermined location in a beam.
This was achieved for both the mono-
lithic and precast concrete specimens.
2. The precast concrete specimens
with the relocated plastic hinges were
comparable in strength and ductility to
cast-in-place specimens where plastic
hinges usually form near the column
face . The relocation of plastic hinges
caused cracking in the beam region,
which is a very desirable characteris-
tic, especially for seismic loading.
3. The precast concrete specimens
with specific connection details used
in this experimental program per-
formed similarly to that of monolithi-
cally cast concrete connections. The
precast concrete connections were
strong enough to force the formation
of a plastic hinge away from the col-
umn face.
4. The critical part of the precast
connections was the welded beam bars
as they initiated the failure of the spec-
imens. Special attention must be paid
to the quality of welding at the precast
plates and reinforcing bars.
5. It was observed that the strains in
the beam stirrups near the column
faces of the precast concrete members
were much greater than that of the
monolithic member. Hence, because
of the discontinuity of the precast con-
crete members, additional shear rein-
forcement is necessary near the beam-
to-column connections in precast con-
crete frames.
6. The shear forces within the con-
nection region of the precast and the
monolithic specimens were almost
identical. It is concluded that the con-
finement reinforcement used within the
joint for monolithic members is ade-
quate for precast concrete members.
7. The intermediate reinforcing bars
in the precast concrete specimens had
less effect on the capacity of the speci-
men early in the test; however, as the
test progressed, these bars contributed
to the specimen capacity.
8. The recommendations of ACI-
ASCE Committee 352 for limiting
joint shear stress and providing con-
fmement for the joint region appear to
PCI JOURNAL
be adequate when applied to precast
concrete frames .
RECOMMENDATIONS FOR
FURTHER RESEARCH
The results of this experimental pro-
gram answered many questions about
precast concrete connections and the
relocation of plastic hinges. Some rec-
ommendations exist for the design of
these specimens, but to develop speci-
fications, additional testing must be
performed on precast concrete connec-
tions with and without relocated plas-
tic hinges.
Future designs of precast concrete
connections should be performed in
such a way as to account for columns
being slightly out of plumb or the
beams being slightly longer or shorter
than specified, as was done on Speci-
mens P2 and PR2. The ability to use
connections which allow for these im-
perfections would make them more
practical and realistic for actual field
construction.
The specimens with the relocated
plastic hinges had much less cracking
within the connection region . It may be
possible to relax the reinforcement re-
quirements within the joints if the relo-
cated plastic hinge concept is used.
Further testing could determine how
these requirements may be reduced.
The use of angles or plates ·caused
the corners of the beam (where con-
nected to the column) to break away. A
possible solution would be to connect
the angles which are welded to the
beam's tension and compression rein-
forcement together with steel bars or
plates to help resist or reduce cracking.
The effectiveness of this and other sim-
ilar approaches must be investigated.
ACKNOWLEDGMENT
This project was sponsored by the
Precast/Prestressed Concrete Institute
under its Research Fellowship pro-
gram. Additional funding was pro-
vided by the College of Engineering
and Mines at the University of Ari-
zona. Reinforcing steel was donated
by the Tanner Companies and Stanley
Septe-mber-October 1993
Structures, both located in Phoenix ,
Arizona. During this study , the first
author was enrolled as a graduate stu-
dent at the University of Arizona. The
authors wish to thank Jeffrey J.
Williamson, former graduate student
at the University of Arizona, for his
assistance in testing the specimens.
REFERENCES
I. ACI-ASCE Committee 352, "Recom-
mendations for Design of Beam-Col-
umn Joints in Monolithic Reinforced
Concrete Structures," ACI Journal,
Proceedings, V. 82, No. 3, May-June
1985, pp. 266-283 .
2. Salmon s, J. R. , " Research Needs
of the Precast Prestressed Concrete
Industry, " PCI JOURNAL, V. 26 ,
No . 6, November-December 1981 ,
pp. 22-30.
3. Hawkins , N. M ., " State-of-the-Art
Report on Seismic Resistance of Pre-
cast Concrete Structures - Part 2:
Precast Concrete," PCI JOURNAL,
V. 23, No. 1, January-February 1978,
pp. 40-58.
4. Pillai, S. U., and Kirk, D. W., "Due-
tile Beam-Column Connection for
Precast Concrete," ACJ Journal, Pro-
ceedings , V. 78, No. 6, November-
December 1981 , pp. 480-487.
5. Martin , L. D. , and Korkosz, W. J.,
"Connections for Precast Prestressed
Concrete Buildings," Precast/Pre-
stressed Concrete Institute, Chicago,
IL, March 1982.
6. Clough, D. P. , "Design of Connec-
tions for Prestressed Precast Concrete
Buildings for the Effects of Earth-
quakes," Technical Report No. 5, Pre-
cast/Prestressed Concrete Institute,
Chicago, lL, March 1985, 144 pp.
7. Bhatt, P., and Kirk, D. W. , "Tests on
Improved Beam-Column Connection
for Precast Concrete," ACI Journal,
Proceedings, V. 82, No. 6, November-
December 1985, pp. 834-843.
8. Dolan, C. W., Stanton, J. F., and An-
derson , R. G ., "Moment Resistant
Connections and Simple Connec-
tions," PCI JOURNAL, V. 32, No. 2,
March-Aprill987, pp. 62-74.
9. Hawkins, N. M., and Englekirk, R. E.,
"U.S.-Japan Seminar on Precast Con-
crete Construction in Seismic Zones,"
PCI JOURNAL, V. 32, No. 2, March-
Apri11987, pp. 75-85.
10. Stanton, J., "Connections in Precast
Concrete Structures," Concrete Inter-
national, V. 9, No. II , November
1987, pp. 49-53.
II. Englekirk, R. E., " An Analytical Ap-
proach to Establishing the Seismic
Resistance Available in Precast Con-
crete Frame Structures," PCI JOUR-
NAL, V. 34, No. 1, January-February
1989, pp. 92-101.
12. French, C. W., Olanrewaju, A., and
Charbel , T., "Connections Between
Precast Elements - Failure Outside
Connection Region," ASCE Journal
of Structural Engineering, V. 115,
No. 2, February 1989, pp. 316-340.
13. Seckin , M., and Fu , H-C ., "Beam-
Column Connections in Precast Rein-
forced Concrete Construction," ACI
Stru ctural Journal, V . 87, No.3,
May-June 1990, pp. 252-261 and dis-
cussions, V. 88, No. 2, March-April
1991 , pp. 240-241.
14. Cheok, G. S. , and Lew, H. S. , "Perfor-
mance of Precast Concrete Beam-to-
Column Connections Subject to
Cyclic Loading ," PCI JOURNAL,
V. 36, No. 3, May-June 1991, pp. 56-67.
15. Cbeok, G. S. , and Lew, H. S., "Model
Precast Concrete Beam-to-Beam Con-
nections Subject to Cyclic Loading,"
PCI JOURNAL, V. 38, No. 4, July-
August 1993, pp. 80-92.
16. Mast , R. F. , " A Precast Concrete
Frame System for Seismic Zone
Four," PCI JOURNAL, V. 37, No. 1,
January-February 1992, pp. 50-64.
17. Priestley , M. J. N. , and Tao, J . R.,
"Seismic Response of Precast Pre-
stressed Concrete Frames With Par-
tially Debonded Tendons," PCI
JOURNAL, V. 38, No. 1, January-
February 1993, pp. 58-69.
18. Standards Association of New
Zealand , "Code of Practice for the
Design of Concrete Structures,"
Wellington, New Zealand, 1982.
19. Abdei-Fattah, B. A., and Wight, J. K.,
"Study of Moving Beam Plastic Hing-
ing Zones for Earthquake-Resistant
Design of RIC Buildings," ACI Struc-
tural Journal, V. 84, No. I, January-
February 1987, pp. 31-39.
20. Al-Haddad, M. S., and Wight, J. K .,
"Relocating Beam Plastic Hinging
Zones for Earthquake Resistant De-
sign of Reinforced Concrete Build-
ings," ACI Structural Journal, Pro-
ceedings, V. 85 , No. 2, March-April
1988, pp. 123-133.
21. ACI Committee 318, "Building Code
Requirements for Reinforced Con-
crete (ACI 318-89)," American Con-
crete Institute, Detroit, MI, 1989.
75