Sunteți pe pagina 1din 13

Model Precast Concrete

Beam-to-Column Connections
Subject to Cyclic Loading
Geraldine S. Cheok Experimental results of eight ~-scale model
Research Structural Engineer
Building and Fire Research precast concrete beam-to-column connections
Laboratory are presented. The test specimens consisted
National Institute of
Standards and Technology
of interior connections designed in accordance
Gaithersburg, Maryland with the 1985 Uniform Building Code pro-
visions for Seismic Zones 2 and 4. These tests
constitute the second and third phases of a
multi-year test program being conducted at the
National Institute of Standards and Technology.
The objective of the test program is to develop
guidelines for an economical precast beam-to-
H. S. Lew, Ph.D. column connection for regions of high seis-
Chief micity. Variables considered in the research
Building and Fire Research
Laboratory program include location of the post-tensioning
National Institute of steel, the use of post-tensioning bars vs. pre-
Standards and Technology
Gaithersburg, Maryland
stressing strands and fully bonded vs. partially
bonded strands. Specimens were subjected to
reversed cyclic loading according to a pre-
scribed displacement history. Comparisons
were made between the behavior of precast
concrete specimens and the monolithic
specimens tested previously in Phase I. These
comparisons were based on connection
strength, ductility and energy dissipation
characteristics. Comparison of results with the
monolithic test specimens indicates that the
post-tensioned precast concrete specimens
had comparable connection strengths, higher
ultimate displacement ductilities and total
energy dissipation to failure, but lower energy
dissipation per cycle.

80 PCI JOURNAL
study of the behavior of YJ-

A scale model precast concrete


beam-to-column connections
subject to cyclic inelastic loading was
D
initiated at the National lnstitute of
Standards and Technology (NIST) in
South ¢=::) North

1987. The objective of the experimen-


tal program is to develop recom- UPPER
CROSS HEAD
mended guidelines for the design of an
economical precast concrete beam-to-
column connection, that provides for 1321
rapid field assembly without the use of
corbels, for regions of high seismicity.
These connections were designed in
accordance with the provisions of the
1985 Uniform Building Code (UBC)' All dimcnsiom in mm
2S.4mm= I inch
for Seismic Zones 2 and 4.

TEST PROGRAM
Fig. 1. Schematic of the loading and boundary conditions for the test specimens.
The test program consists of four
phases. Phase I was an exploratory
phase in which four monolithic speci-
mens and two precast concrete speci-
mens were tested. Test results of the
monolithic specimens served as refer-
ences for the precast concrete tests.
The results from the precast tests were
used to determine the viability of the
connection details.
Phase II of the program involved
testing six precast concrete specimens.
As a result of the low energy dissipa-
tion observed for the precast concrete
specimens in Phase I, several methods
of increasing the energy dissipation of
the precast concrete connection were
explored.
Because of a lack of stiffness in the
precast specimens in the later stages of
the tests upon load reversal, the use of
partially bonded tendons was studied Fig. 2 . Test setup and loading arrangement.
in Phase III. Phase IV, currently un-
derway, involves the study of using
and loading arrangement are shown in defined as the average of these two
conventional mild reinforcing steel in
Fig. 2. displacements divided by 0.75. As de-
conjunction with a post-tensioned pre-
An axial load equal to 0.1 f~ Ag was fined in the previous paragraph, the
cast connection.
applied to the column at the beginning yield displacement obtained for the
The results of Phases II and III of
of the test and maintained throughout precast concrete specimens was actu-
the NIST test program are summarized
the test. The column was then laterally ally a nominal yield displacement as
in this paper. Results from Phase I and
loaded at the column top for the initial yielding of the post-tensioning steel
Phase II were reported by Cheok and
elastic cycle. This initial elastic cycle did not occur in the initial elastic cycle.
Lew in Refs. 2, 3 and 4.
developed a load equal to 75 percent Henceforth, the term yield displace-
of the maximum strength of the mono- ment should be taken to mean the
Test Procedure lithic beam calculated using nominal nominal yield displacement.
Boundary conditions for the test values for the material properties. The After the initial elastic cycle, the
specimens are shown schematically in top column displacements in the for- basic loading history was two cycles
Fig. 1. The column bottom was pinned ward (south) and reverse (north) direc- each at ± 2L\y and ± 4L\y, three cycles
and the column top and beam ends tions were measured. each at ± 6L\y and ± 8L\y, two cycles at
were roller supported. The test setup The "yield" displacement, L\y, was ± 1OL\y, and three cycles at ± l2L\y.
July-August 1993 81
Table 1. Specimen description.

Bar distance from Length of debonded


Post-tensioning extreme fiber post-tensioning steel
Test phase Specimens Seismic zone Type* steelt Groutedt (mm) (mm)

I A-M-Z2 & B-M-Z2 2 M - - - -

I A-M-Z4 & B-M-Z4 4 M - - - -

I A-P-Z4 & B-P-Z4 4 p B F 89 -

II A-P-Z2 & B-P-Z2 2 p s F 63 -

II C-P-Z4 & D-P-Z4 4 p B F 135 -

II E-P-Z4 & F-P-Z4 4 p s F 102 -

III G-P-Z4 & H-P-Z4 4 p s p 102 1219

* M = Monolithic; P = Precast
B =Post-tensioning bars; S =Prestressing strands
F =Fully grouted; P =Partially grouted
Note: 25.4 mm = I in.

This loading history is similar to those design guidelines were as strong and as bonded through the joint region.
used by other researchers [e.g., Bhatt ductile as the monolithic Zone 4 speci- This concept was developed by
·and Kirk, 5 Bull and Park, 6 French et mens. However, compared with the Priestley and Tao. 8 Briefly, the strains in
al. 7 ]. The test was stopped whenever monolithic specimens, the cumulative the post-tensioning steel would be kept
failure occurred. and average cyclic energy dissipated by in the elastic range and the post-tension-
Displacement ductility, )1, is defined the precast concrete specimens were 20 ing force would, therefore, be main-
as the ratio of the maximum displace- and 70 percent lower, respectively. tained at higher drift levels and thereby
ment achieved in any cycle to the In an effort to improve the energy preserve the connection stiffness upon
yield displacement. Ultimate displace- dissipation characteristics of the pre- load reversal. Special interlocking spi-
ment ductility, Jlu, is defined as the cast concrete connection, several pa- rals were provided in the beam hinge
ratio of the maximum displacement rameters were studied in Phases II and region because high compressive strains
achieved at failure to the yield dis- III. One such parameter was the loca- were expected in this location. A volu-
placement. Failure was considered to tion of the post-tensioning steel in the metric ratio of 2 percent with a maxi-
have occurred when the lateral force beams. It was felt that the integrity of mum spiral pitch of D/4, where D is the
during a cycle dropped below 80 per- the connection would be improved if spiral diameter, was used to design the
cent of the maximum lateral load that the post-tensioning steel were moved confining spirals.
was achieved in the first cycle at 2~y· closer to the beam centroid. By doing A summary of the specimens tested
Strains in the beam steel and ties this, the post-tensioning steel would in Phases I, II and III is given in
were measured using resistance-type experience less strain and would, Table 1. Phase II of the test program
strain gauges. The gauges were placed therefore, retain its clamping force at consisted of testing three sets of speci-
at specified locations along the beam higher drift levels. mens. The Phases II and III precast
flexural steel. In the case of the pre- Also, the difference in connection specimens were designed in a manner
cast concrete specimens, only the mild behavior by the use of prestressing similar to the Phase I specimens. The
steel reinforcing bars in the beam were strands instead of post-tensioning bars required amount of post-tensioning
instrumented (not the post-tensioning was examined. This variable was in- steel was computed so that the pre-
steel). In addition, the applied lateral cluded in the study because prestress- dicted strength of the precast connec-
load at the column top, the displace- ing strands are used more frequently tion was the same as that of the mono-
ment of the column top, beam loads in practice. lithic connection.
and concrete strain at various locations When the Phases I and II tests were Reinforcement details for the precast
along the beam were measured conducted, concern was raised regard- columns, for both Zones 2 and 4, were
throughout the tests. ing the zero slope of the hysteresis loops the same as those of the corresponding
upon load reversal that was observed in monolithic specimens. The reinforce-
the precast concrete tests. This lack of ment details for the Zone 4 precast
Phases II and Ill Specimens
stiffness observed in the later stages of beams were identical to those of the
The results of the Phase I tests indi- the tests was thought to be caused by Phase I precast specimens except for
cated that post-tensioned, precast con- the inelastic response of the post-ten- the location of the post-tensioning
crete specimens designed using the sioning steel. A possible solution to this steel.
1985 UBC Seismic Zone 4 criteria as problem was to leave the strands un- Fig. 3 shows the reinforcement de-

82 PCIJOUANAL
2-#3
Top and B«tan 76
140

+
2Smm
Pcst-k:Mioning bar 76
&
38mm
COI'l'UJated duct
126 406 406 102
Smm
smlh. wire @
j_
68mm0.C.
3-llmm
Strands
Top&Bottom

2-#3
All diJDimsi0111 in 111m Top and Bottom
25.4 111m = 1 inch
Specimens E & F - Fully Bonded
Specimens C & D
Specimens G & H - Partially Bonded

Fig. 3. Beam cross sections for Phases II and Ill precast concrete Zone 4 specimens.

tails for the precast Zone 4 specimens Table 2. Load cell readings after post-tensioning.
and Fig. 4 shows the reinforcement
Specimen Load Celli* Load Cell2t
details for the Zone 2 specimens. The
reinforcing bars located in the corners E-P-Z4 0.65 F;,u 0.65 F;,u
of the beams did not cross the joint
F-P-Z4 0.65 F;,u 0.67 F;,u
since they were used mainly to hold
the ties together. The main resistance G-P-Z4 0.66 F;,ut 0.66 F;,ut
to the applied loads was provided by
H-P-Z4 0.68 F;,u:f: 0.64 F;,u:f:
the post-tensioning steel.
A-P-Z2 0.71 F;,u§ 0.67 F;,u§
Material Properties B-P-Z2 0.69 F;,u§ 0.70 F;,u§
The stress-strain curves for the #3 re-
* Top set of strands.
inforcing bars, smooth wire used in the Bottom set of strands.
beams and the high strength steel are Average of loads from strain gauges and load cells.
shown in Figs. 5 through 12. The stress- 13 mm (0.5 in.) prestressing strands.

strain curves for the smooth wire used


in Specimens A-P-Z2 and B-P-Z2 and
used - one top and one bottom for day after the post-tensioning operation
E-P-Z4 and F-P-Z4 are similar to those
Specimens E-P-Z4 and F-P-Z4 and A- and the time between post-tensioning
for Specimens G-P-Z4 and H-P-Z4.
P-Z2 and B-P-Z2. Six load cells were and testing the specimens ranged from
used for Specimens G-P-Z4 and H-P- 2 to 4 weeks.
Application of Z4. In addition to the load cells, the
Post-Tensioning strands in Specimens G-P-Z4 and H-
A major concern in the use of P-Z4 were instrumented with strain TEST RESULTS
strands was the load loss in the strand gauges. These gauges were located in
forces due to seating of the wedges. the unbonded part of the strands. Load-Displacement and Ductility
This loss was expected to be signifi- The loads measured immediately The hysteresis loops for Phases II
cant due to the short lengths of strands after post-tensioning are given in and III specimens are shown in Figs.
involved. A procedure which involved Table 2. As indicated in Table 2, the 13 through 20. The load and displace-
shimming of the chucks was followed loads in the individual load cells were ment in these figures represents the
to minimize the load loss. within one standard deviation of the applied lateral force and the displace-
When the test specimens were post- mean as obtained in the trial runs. The ment at the column top. The hysteresis
tensioned, only two load cells were post-tensioning steel was grouted one plots for the Zones 2 and 4 monolithic

July-August 1993 83
11691
2-#3
Top and Bottom
~ ~ -...-----..---- r- 220----1 4 -#3,3 -#4
63 Top and Bottom

2 - 13 nun strands 0
_j_

5 nun smth. wire 5nun


@59nunO.C. 254 smth. wire
@51nunO.C.

2-llnun
0
strands
t 63 All mmensions in mm

25.4 mm = 1 inch

BEAM CROSS SECTION COLUMN CROSS SECTION

Fig. 4. Reinforcement details for Phase II precast concrete Zone 2 specimens.

600 .-------,,...------,-1~----,-~--=,...,...-----, 87.0


500 r----;--;----,.----,---r-L---,-----..., 72.5
400 f.=i===t==t::=:=t:::::::::::::. j.................. i... .... j 58.0 500 ······7F···=···-+---f·······-···-··-·-·+·······-····-·······+······················ ·I 72.5

~ ~
~
300 t···l·············+················+··-··········+:················t·················!·-···············!·····.. ·········1 43.5

~0
g
I
~
1
~g : ~- ·.- .-,l. . . . . . . . . . f···········-·············f···························f···-···..··················i..················. . ·······l : : I
~

100 ! ! 14.5
~ 200
~v
u/. . . . . . . . . . . .;. . . . . . . . . . . . . .!·························+························+···················· I ~ 29.0

[
0 '---"'----'----'---'---'---'--.....1 0.0 0 ' - - - - ' - - - - ' - - - - - - ' ' - - - - - - ' ' - - - - - - - ' 0.0
0 0.005 O.QJ O.Q15 0.02 0.025 0.03 0.035
0 0.005 0.01 O.oJ5 0.02 0.025
Strain
Strain

Fig. 5. Stress-strain curve for #3 reinforcing bars, Fig. 6. Stress-strain curve for #3 reinforcing bars, all
Specimens A-P-Z4 to D-P-Z4, A-P-Z2 and B-P-Z2. monolithic, Specimens E-P-Z4 and F-P-Z4.

specimens (from Phase I) are shown in dons and debonding of the strands ments and ultimate displacement duc-
Figs. 21 and 22, respectively, for com- during the test which allowed some tilities for all the specimens (Phases I,
parison purposes. slippage to occur. II and III) are listed in Table 3. It
The ductilities of the precast Zone 4 The shape of the hysteresis loops should be noted that the yield displace-
specimens, as shown in Figs. 13 for the fully bonded precast specimens ments for the precast concrete speci-
through 18, are greater than that of the (see Figs. 15 and 16) differ from those mens in Table 3 are nominal yield dis-
monolithic Zone 4 specimen, as can be for the partially bonded specimens placements and J.lu is based on this
seen from Fig. 22. However, the hys- shown in Figs. 17 and 18. As seen in nominal yield displacement. Also listed
teresis loops for the precast concrete these figures, the partially bonded in Table 3 are the initial connection
specimens are more severely pinched specimens did not exhibit zero stiff- stiffnesses and story drifts at failure.
than those for the monolithic speci- ness upon load reversal. However, the Both precast Zone 2 specimens had
men. The pinched hysteresis loops loops for the partially bonded speci- an ultimate displacement ductility of
probably result from the combination mens are narrower than those for the 4, which is 50 percent less than the
of the concentration of rotation at the fully bonded specimens. ductility of the corresponding mono-
column face due to yielding of the ten- The experimental yield displace- lithic specimens. The yield displace-

84 PCI JOURNAL
600 87.0 600 87.0

L-- -···-·····_[
----
500 72.5 500 72.5

'2 400 . .7 58.0 400


I 58.0
ll..
g ~ 1i
~
6 43.5 6 ·:::::::-; 43.5
300 300
~
"'
!l"'
<1.1
200 :::.J 29.0
"'
g
<1.1
"'
!l"'
<1.1 200 I .....J 29.0 ":'

100 1./
1/
14.5 100 I
1/
14.5

0 i 0
0 0
0 0.005 O.QJ O.Ql5 0.02 0.025 0 0.004 0.008 0.012 0.016 0.02
Strain Strain

Fig. 7. Stress-strain curve for #3 reinforcing bars, Fig. 8. Stress-strain curve for smooth wire, Specimens
Specimens G-P-Z4 and H-P-Z4. A-P-Z4 to D-P-Z4.

ments for the Zone 2 specimens, tests for these specimens were stopped increases in stiffness were 85 percent
monolithic and precast, were approxi- prior to failure. At Jlu = 14 (story drift for precast Specimens A, B, G and H,
mately the same. However, due to the ""4 percent), the specimens did not ex- 95 percent for precast Specimens C
higher ductilities achieved by the hibit any signs of strength degrada- and D, and 135 percent for precast
monolithic specimens, the story drifts tion, and it was thought that the speci- Specimens E and F. The fully bonded
at failure were 65 percent greater for mens would not fail unless the strands precast specimens post-tensioned with
the monolithic specimens. fractured. The expected displacement strands (Specimens E and F) were
As listed in Table 3, all the Zone 4 ductility of 15.6 for these specimens 8 stiffer and achieved slightly higher
precast specimens achieved higher ul- compares well with the experimental story drifts at failure than did the spec-
timate displacement ductilities than minimum value of 14. imens post-tensioned with fully
those of the monolithic specimens. The initial connection stiffnesses bonded bars (Specimens C and D).
The displacement ductilities listed in for the Zone 4 precast specimens were Also, the precast specimens with fully
Table 3 for Specimens G-P-Z4 and H- higher than those for the companion bonded steel had slightly higher initial
P-Z4 are minimum values because the monolithic specimens. Approximate connection stiffnesses when the post-

Table 3. Yield displacement and ultimate displacement ductility.

Concrete compressive Experimental yield Ultimate Initial elastic connection Ultimate story
Specimen strength, J; * displacement displacement stiffnesst drift
designation MPa (psi) mm (in.) ductility, 11u kN/m (kips/in.) (percent)

A-M-Z2 43.5 (6310) 9.1 (0.359) 6 7.7 (44) 4.1


B-M-Z2 41.1 (5960) 9.4 (0.371) 6 6.6 (38) 4.3

A-P-Z2 34.0 (4930) 8.5 (0.333) 4 12.4 (71) 2.6


B-P-Z2 36.4 (5280) 8.4 (0.330) 4 11.0 (63) 2.5

A-M-Z4 30.7 (4450) 6.7 (0.263) 6 21.2 (121) 3.0


B-M-Z4 32.2 (4670) 7.4 (0.293) 6 18.0 (103) 3.4

A-P-Z4 40.6 (5890) 4.1 (0.160) 10 35.7 (204) 3.1


B-P-Z4 44.5 (6450) 4.5 (0.179) 10 37.8 (216) 3.4
C-P-Z4 46.8 (6780) 5.3 (0.209) 12 31.5 (180) 4.8
D-P-Z4 44.9 (6510) 5.4 (0.213) 12 45.2 (258) 4.9
E-P-Z4 29.2 (4230) 5.7 (0.225) 12 42.2 (241) 5.2
F-P-Z4 27.7 (4020) 5.5 (0.218) 12 50.1 (286) 5.0
G-P-Z4 30.1 (4370) 3.6 (0.143) 14:j: 36.1 (206) 3.9
H-P-Z4 32.2 (4680) 3.4 (0.135) 14:j: 35.4 (202) 3.6

* Strengths were obtained at the time of specimen testing.


These values are the initial elastic stiffness as obtained from the slope ofihe load-displacement plot for the initial excursion to +0.756,.
The specimens did not fail at this ductility level. Therefore, these are minimum values for the specimens.

July-August 1993 85
600 r-----,,..------,----;----;---, 87.0 2600 290

soo400 ~:::::--;r-=f=. =+==:


. . . . T=t==+==ln.s 1600 232

... ;+·--·········!···· ················-'·······-········· ·····+·····················+--············· ··-l 58.0

~ ~"'
l
6
1200 174 g..
R
<ll
300

I
-!-··-···········+········-·············+············-········+···················· +··-···················! 43.5

200 't·-,/···············+-···········-·······+-··-··---····-+·······-··-·········+·············-········1 2!1.0


g
<ll
::g
.,t lnl 116
g
<ll

400 ····---- 51
100 ll!l··············-·····+······················-+······-········--··-t·······················+·······················l14.5

0 0 0 0
0 0.005 O.Ql O.oi5 0.02 0.025 0 o.oos 0.&1 0.015 0.02 0.025
Strain Slr.tin

Fig. 9. Stress-strain curve for smooth wire, Specimens Fig. 10. Stress-strain curve for 7/16 in. (11 mm) strands,
G-P-Z4 and H-P-Z4. Specimens E-P-Z4 to H-P-Z4.

tensioning steel was moved closer to material strengths and strain hardening. The measured maximum moments
the beam centroid. No factor for strain hardening was used for the monolithic specimens were
to calculate the moments for the pre- achieved at or close to the ultimate dis-
cast concrete specimens. placement ductility of the specimens.
Flexural Strength The moments for the partially This was also true for precast Speci-
The maximum measured and calcu- bonded specimens were calculated mens A-P-Z2, B-P-Z2, A-P-Z4 and B-
lated moments are listed in Table 4. based on ACI-318 Eq. (18-4) 9 for de- P-Z4. However, the measured maxi-
The calculated values were based on an termining the stress in the strands at mum moments for the precast concrete
ultimate concrete compressive strain of nominal strength. Again, actual mate- specimens with the post-tensioning
0.003 and actual material properties. rial properties were used. The experi- steel closer to the beam centroid (Spec-
The calculated ultimate moments for mental values were obtained by multi- imens C-P-Z4 through F-P-Z4) were
the monolithic specimens had a factor plying the peak load recorded in the achieved earlier in the tests (at J..l "' 4
of 1.25 applied to the steel yield stress beam load cell and the moment arm to whereas J..lu "' 12). The measured maxi-
to account for possible higher actual the column face. mum moments for the partially bonded

Table 4. Experimental and calculated maximum moments.

Concrete compressive strength Calculated maximum Experimelltal maximum


Specimen J;* moment momentt Avg. exp. moment
designation (2) (3) (4) Calc. max. moment
(I) MPa (psi) kN-m kip-ft kN·m kip·ft (5)

A-M-Z2 43.5 (6310) 68 50 69 & 80 51 &59 1.10


B-M-Z2 41.1 (5960) 68 50 70& 75 52& 55 1.07

A-P-Z2 34.0 (4930) 46 34 52 &54 38 &40 1.15


B-P-Z2 36.4 (5280) 46 34 50&54 37 &40 1.13

A-M-Z4 30.7 (4450) 132 97 148 & 144 109&106 1.11


B-M-Z4 32.2 (4670) 132 97 148 & 153 109 & 113 1.14

A-P-Z4 40.6 (5890) 155 114 176 & 186 130 & 137 1.16
B-P-Z4 44.5 (6450) 155 114 184 & 186 136 & 137 1.20
C-P-Z4 46.8 (6780) 145 107 169 & 171 125 & 126 1.17
D-P-Z4 44.9 (6510) 145 107 165 & 169 122 & 125 1.15
E-P-Z4 29.2 (4230) 111 82 138 & 136 102 & 100 1.24
F-P-Z4 27.7 (4020) 111 82 146 & 137 108 & 101 1.28
G-P-Z4 30.1 (4370) 95 70 118 & 123 87 &91 1.27
H-P-Z4 32.2 (4680) 95 70 130 & 132 96&97 1.38

* Strengths were obtained at the time of specimen testing.


t Moments are at the two column faces.

86 PC! JOURNAL
~ 290

I~
1:!00 174

1600 232 1000 145

t:
6
1200 I 174 S'
.
~ l 100 ............ / 116 ~
·;;

Il"'
/ ~
6
1'1
600
···-·.../ 87
~

g
800 116
"' g
.)
"'
400
7 58
"' 400 58 "'

0
0
17 0.004 0.008 0.012 0.016 0.02
0
200

0
If.. 29

0
0 0.005 O.oJ O.oJ5 0.02 0.025 0.03 0.035
Strain Sttain

Fig. 11. Stress-strain curve for 1/2 in. (13 mm) strands, Fig. 12. Stress-strain curve for 1 in. (25 mm) Dywidag bars,
Specimens A-P-Z2 and B-P-Z2. Specimens A-P-Z4 to D-P-Z4.

specimens (minimum J.lu = 14) were the results of Specimens A-P-Z4 and B- than that of the fully bonded speci-
obtained at}.l"" 10. P-Z4 with those of Specimens C-P-Z4 mens (E-P-Z4 and F-P-Z4).
It would appear that the fully and D-P-Z4, the reduction in strength
bonded precast concrete specimens caused by moving the post-tensioning
with the post-tensioning steel located bars closer to the beam centroid was ap- Failure Modes
closer to the beam centroid debonded proximately 8 percent. The expected re- The failure mode for the precast
early in the test and behaved in a man- duction in connection strength (Column Zone 2 specimens was beam crushing.
ner similar to the partially bonded 3, Table 4) for these specimens was 6 This is different from the failure mode
specimens. Debonding of the steel in percent. The ratios given in Column 5 for the monolithic Zone 2 specimens,
the specimens with the post-tensioning of Table 4 indicate that the precast which failed predominantly in shear in
steel further away from the beam cen- concrete specimens post-tensioned the column joint region. The precast
troid may also have occurred, and the with strands had more "reserve" specimens did not sustain significant
higher moments achieved in the later strength than those post-tensioned damage in the joint region. The beam-
stages of the tests may be a result of with high strength bars - an average to-column opening for the precast
the strain hardening of the steel. of 29 percent greater than the calcu- Zone 2 specimens was approximately
As shown in Column 4 of Table 4, lated moment for the strands as op- 5 mm (0.2 in.). This is one-fifth of the
the fully bonded precast Zone 4 speci- posed to 17 percent for the bars. The opening size for the precast Zone 4
mens (A-P-Z4 to F-P-Z4) were as experimental strength of the partially specimens. There are insufficient data
strong as or stronger than their mono- bonded specimens (G-P-Z4 and H-P- to determine if this is a result of the
lithic counterparts. When comparing Z4) was approximately 10 percent less higher initial beam stress resulting

Displacement (in.) Displacement (ifl.)


-2 -1 0 2 -2 -1 0 2
~ 90 ~ .-----.-----~----~----. 90
Fully Bonded Bars l: .Fully Bonded Bars 67
67 300

: .::::::::::::::::J::::::::::::::::::r····.
: : 4S 200 :::::::::::::::::::r::::::::::::::::r······· 4S

100 ···················l···················lflb'Hn'H¥-(I#.flf 22 100


! !
•••••u•••••••••••!•••••••••••••••••••i 22
~ 0 . : g ~ 0

~ -100 ~fj jliJI I.fl l· · · · ; e:· ·:;.;· · ; c· ·~-+-"'· · · · · · · · ·


: :
-22 ~ ~ -100
: i
't"""'""""""t"""""""""' -22

-200
··!·················+················· -4S -200
:
:
......t.................. t". . . . . . . .
:
:
-45
: :
-300 ·········-·--------i---······· ······-i---·--------------~---·-········ ..··· -67 -300 ...................;.................i..................~.................. -67

~ ~
: : !
i 1 i
-~ L-----~----~----~----~ -90 -~
l -90
-100 -SO 0 100 so -100 -SO 0 SO 100
DisplacemeRt (mm) Displacement (mm)

Fig. 13. Hysteresis plot for Specimen C-P-Z4. Fig. 14. Hysteresis plot for Specimen D-P-Z4.

J!My-August 1993 87
Displacement (in.) Displacement (in.)
-2 -1 0 2 -2 -1 0 2
«0
«0 r-----,1----F-u~ll-y_B_o_n~~-S-tr_m_&-, 90 90

67 300 67
: .::::::::::::::::::l:::::::::::::::::::r:················t·················· 45 200 45
: :
~ 100 ···················j···················[_.Afr#-7/ 22
g 100 22 g
~
~ ~
0 0 0 0

~ -100 i--······················-··········.. -22 ~ -100 -22

-200
:
..\..................1..................
:
-45 -200 -45
i ! -300
-300 ••••••••••••••••••• f.................·-i·················-+·················· -67 -67

-400
~ j i
L-----~----~----~--~ -90 -«0 -90
-100 -SO 0 50 100 -100 -50 0 50 100
Displacement (mm) Displacement (mm)

Fig. 15. Hysteresis plot for Specimen E-P-Z4. Fig. 16. Hysteresis plot for Specimen F-P-Z4.

from the post-tensioning of the Zone 2 umn, approximately 50 percent greater, Energy Dissipation
specimens, or the result of the lower was observed in the precast specimens Characteristics
loads experienced by these specimens. which had the post-tensioning steel Fig. 24 shows a comparison of the
The failure modes for all the precast closer to the beam centroid. energy dissipated per cycle for all the
Zone 4 specimens were similar to The partially bonded precast speci- precast Zone 4 specimens. As seen in
those for the precast Zone 2 speci- mens experienced more crushing of the chart, Specimens E-P-Z4 and F-P-
mens. One of the precast Zone 4 con- the beams than did the fully bonded Z4 (fully bonded strands) performed
nections after failure is shown in specimens. However, the width of the the best and Specimens G-P-Z4 and
Fig. 23. The precast concrete beams opening between the beam and col- H-P-Z4 (partially bonded strands)
sustained less cracking than did the umn was similar in both sets of precast performed the worst in terms of en-
beams in the monolithic specimens, concrete specimens. The column joint ergy dissipated per cycle. Again, the
and concentration of beam hinging at regions in the partially bonded precast low cyclic energy dissipated by the
the column face was observed. How- specimens appeared to sustain the partially bonded specimens was due
ever, more beam crushing and a wider same amount of damage as the fully to the expected bilinear elastic behav-
opening between the beam and col- bonded set of precast specimens. ior of these specimens.

Displacement (in.) Displacement (in.)


-2 -1 0 2 -2 -1 0 2
«0 r--.--.-~--~~--~~--. 90 400 90

300 67 300 67
200 45 200 45
100 22 100
g 22
~ ~ g
~
0 0 0
~
0
~ -100 -22 ~ -100 -22
-200 -45 -200 -45
-300 -67 -300 -67
-400 -90 -400 -90
-80 -60 -40 -20 0 20 40 60 80 -80 -60 -40 -20 0 20 40 60 80
Displacement (mm) Displacement (mm)

Fig. 17. Hysteresis plot for Specimen G-P-Z4. Fig. 18. Hysteresis plot for Specimen H-P-Z4.

88 PCI JOURNAL
Displacement (in.) Displacement (in.)
-2 -1 0 2 -2 -1 0 2
~ r----.----~--~--~~--~
:
90
400 ~------------------,---~ 90
! Fully Bonded Strands i Fully Bonded Strands

~-~~-[~-~-
67 300 ··············t··············t··············t··············;··············· 67
: : : :
: l : !
4S 200 ··············;··············t··············t··-··········!··············· 45
i ! 1 i

I _:~ =:=:-gz~::_:=:
22
g
22 g
0
~
0 ]
: ! l :
-22 -22
! i ! l
-200 ..............J..............l..............l-..............1............... -45 -200 ···············~·-············t··············t··············i··············· -45
-300

-400
-100
---+--+--+--+----
-60 -20 20 60 100
-67

-90 -400
! !
!:

l ~
L---~----~--~----~--~
-100 -60
i:

-20
i:

20
i:
-300 ···············r·············r··············r··············1···············

60 100
-67

-90

Displacement (mm) Displacement (mm)

Fig. 19. Hysteresis plot for Specimen A-P-Z2. Fig. 20. Hysteresis plot for Specimen B-P-Z2.

Comparing the results of Specimens post-tensioning bars (Specimens C-P- H-P-Z4 was less than the cyclic en-
A-P-Z4 and B-P-Z4 with those of Z4 and D-P-Z4). As seen by the large ergy dissipated by the monolithic
Specimens C-P-Z4 and D-P-Z4, the standard deviations, the percent in- Zones 2 and 4 specimens. This is due
energy dissipated per cycle showed an crease in energy dissipated per cycle to the narrow and pinched hysteresis
average increase of 45 percent ( cr = 24 was highly variable. The energy dissi- loops for these specimens.
percent) when the post-tensioning bars pated per cycle decreased by approxi- The cumulative energy dissipated
were moved closer to the beam cen- mately 55 percent (cr = 7 percent) to failure by all the specimens is
troid. This increase is probably due to when partially bonded strands were shown in Fig. 26. With respect to cu-
the increased crushing of the beam used instead of fully bonded strands. mulative energy dissipated, the precast
concrete in Specimens C-P-Z4 and D- As shown in Fig. 25, the energy dis- Zone 4 Specimens C through F per-
P-Z4 as observed during the tests. sipated per cycle by precast Speci- formed better than the monolithic
An average increase of 30 percent mens E-P-Z4 and F-P-Z4 was an aver- Zone 4 specimens. This was a result of
( cr = 22 percent) in the energy dissi- age of approximately 60 percent of the higher displacement ductilities
pated per cycle was noted when pre- that dissipated by the monolithic Zone achieved by these precast concrete
stressing strands (Specimens E-P-Z4 4 specimens. Also, the cyclic energy specimens, which meant that these
and F-P-Z4) were used in place of dissipated by Specimens G-P-Z4 and specimens underwent more reversed

Displacement (in.) Displacement (in.)


-2 -1 0 2 -2 -1 0 2
400 ,----,----,.----,.----.,---~ 90 400 90

·············r············l-···-~~~r~~--1~~~--~---·
300
: ! Mon~lithic..···:z:one
··············-:·······..·····-:··········.. ..
2
·-~---···· ~·-············· 67
i: !: i: i: 300 67
200 ···----····----:-·-······-····-=-·········-····+··············~·-············· 45 200 ··············~·············+·········· ..... ··i··············· 45
: : :

~ ···············!~············+············! 22
i i
100 ··············-'··············~·····.. ..~
i
.... ·····•··············· 22
g
100 .. ·····!··············· ~ 0 ...............\......... : .... ' ...........1............... g
~
0 ···············\···· ..... . ..............j............... 0 0
~
~ ~
: : :
: : : ! ~ !
:
!
: .....l
·1 00 ···············!········ . :··············t··············i··············· -22 -100 ···············r···· .. ~ ...... ............... ............... -22
!
!
!
:
:
:
:
l !
1: 1:
...............j. ••••••••••••••.t .............. .(. .............. ~·-·············
-200 -45 ··············r·· .... .. . ··········r-··········l·············
-300 ...............1..............
:
i
!. . . . . . . !. . . . . . . 1...............
:
:
:
:
:
:
-67
-200

-300 ··············r············r···········r··········l·············
-45

-67
-400 L---~·----·~--~·~--~·~~ -90 -400 -90
-100 -60 -20 20 60 100 -100 -60 -20 20 60 100
Displacement (mm) Displacement (mm)

Fig. 21. Hysteresis plot for Specimen A-M-Z2. Fig. 22. Hysteresis plot for Specimen A-M-Z4.

July-August 1993 89
specimens. This is due to the concen-
tration of the rotation at the column
face for the precast concrete speci-
mens. The recorded strains in ties for
the precast Zone 4 specimens were
also much lower than for the mono-
lithic specimens. Among the precast
specimens, the specimens post-ten-
sioned with bars had longer reinforc-
ing bar yield lengths than those post-
tensioned with strands. It would
appear that the strains and rotations
were more concentrated at the column
joint for the specimens post-tensioned
with strands.

CONCLUSIONS
Fig. 23. Representative failure mode for the precast concrete specimens. 1. Failure Mode - Failure modes
for all precast concrete specimens were
similar. Failure was characterized by
yielding of the post-tensioning steel,
16 .---------------.------------------------, 142 beam crushing, and an opening at the
Ill Avg. A & B P-Z4 junction between the beam and the col-
D Avg.C&DP-Z4
Ill Avg. E & F P-Z4
umn. The width of the opening at the
1:-:s Avg. G&HP-Z4 beam-to-column joint increased as the
12 \---------------'················································· ............... ...... ··········· 106
post-tensioning steel was placed closer
to the beam centroid. The width of the
opening does not appear to be influ-
8 .................................................. ...................... ...... 71 enced by the type of post-tensioning
steel (post-tensioning bars or prestress-
ing strands) used to connect the precast
4 ............................... 35
concrete elements or by the use of par-
tially debonded prestressing. It should

I
be noted, however, that the partially
bonded specimens were not tested to
0 0 failure.
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Unlike the monolithic Zone 2 speci-
Cycle Number mens which failed predominantly in
shear in the column joint region, the
Fig. 24. Cyclic energy dissipated by precast Zone 4 specimens. precast Zone 2 specimens did not ex-
perience severe joint distress.
2. Displacement Ductility - The
cyclic loading prior to failure. therefore, an indication of the concen- ultimate displacement ductility for
The cumulative energy dissipated to tration of the beam rotation. The yield both precast Zone 2 specimens, f.lu = 4,
failure by Specimens G-P-Z4 and H-P- length of the reinforcement was mea- was lower than that obtained for the
Z4 was approximately equal to that for sured from the column face. This monolithic Zone 2 specimens, f-Lu = 6,
monolithic Zone 4 specimens. How- length was obtained by first plotting and as a result, the precast concrete
ever, these specimens had not failed the strain in the reinforcing bar against specimens had lower story drifts at
when the tests were stopped. There- the distance of the strain gauge from failure.
fore, the cumulative energy dissipated the column face. The points were then In general, the precast concrete
by Specimens G-P-Z4 and H-P-Z4 as connected with straight lines . The specimens had higher story drifts at
shown in Fig. 26 are minimum values. yield lengths were defined as the failure and higher initial stiffnesses
lengths from the column face to the than the monolithic specimens. The
point at which these lines crossed the ultimate displacement ductilities of the
Reinforcing Bar Yield nominal yield strain level. precast specimens (J.lu = 12 and 14)
The yield lengths of the beam rein- As expected, the monolithic speci- were greater than those of the mono-
forcement give an indication of the mens exhibited longer reinforcing bar lithic Zone 4 specimens, (J.lu = 6). The
strain concentration in the beam and, yield length s than did the precast experimental value of 14 for the par-

90 PCIJOURNAL
dissipated per cycle by Specimens E-
20 r---------------.------------------------. 177 P-Z4 and F-P-Z4 (the precast speci-
IIIII Avg.A&BM-Z2
r.d Avg. A & B M-Z4 mens which performed best with re-
IIIII Avg.A&BP-Z4 spect to energy dissipated per cycle)
~
15 8 Avg.C&DP-Z4
~:::~:~~~ ............................................. 133 was, on the average, approximately 60
percent of tllat for tlle monolithic Zone
4 specimens.
In general, the cumulative energy
10 .............................................................................................................................................. 89
dissipated to failure by precast Zone 4
specimens was greater than the cumula-
tive energy dissipated by the monolithic
5 .............................................................. 44 Zone 4 specimens. This is due to the
higher displacement ductilities achieved
by the precast concrete specimens.

0
2 3 4 5 6 7 8 RECOMMENDATIONS
Cycle Number Test results obtained thus far show
that a post-tensioned, precast beam-to-
Fig. 25. Comparison of energy dissipated per cycle to 6~y (Cycle 3). column connection appears to be a vi-
able system for regions of high seismic
activity. The shear friction between a
precast beam and column has been
l4U 103
shown to be sufficient to resist tlle ap-
plied shear loads witllout the need for
120 89
corbels or shear keys. This type of
post-tensioned precast concrete con-
100 74
nection has also behaved as well as,
8 ;E' and in most cases, better than similar
~ 80 59
~ monolithic connections in terms of

~ 60 44 ~ connection strength and ductility.


In regions where energy dissipation
&i &i
30 is not an issue, the partially bonded
40
specimens would be an attractive alter-
15 native as tlley have been shown to per-
20
form in what is essentially an elastic
0
manner. Also, the amount of spiral
0
2 3 4 5 6 7 8 9 10 11 12 13 14 steel used in the beam, with a volumet-
ric ratio of 2 percent and a maximum
Specimen Number pitch of D/4, was sufficient to confine
the beam in the plastic hinge region.
Fig. 26. Comparison of the cumulative energy dissipated to failure.
Note: Specimens G-P-Z4 and H-P-Z4 had not failed when testing was stopped.
FUTURE
tially bonded specimens compared ments. Placement of the post-tension- RESEARCH NEEDS
well with the predicted value of 15.6 ing bars closer to the beam centroid The precast concrete connections
by Priestley and Tao.• does not appear to have a significant exhibited low energy dissipation per
3. Connection Strength - The adverse effect on the connection cycle as compared with monolithic
measured maximum connection strength. connections. Phase IV of the test pro-
strengths for all the precast concrete 4. Energy Dissipation - As in gram will examine the use of low
specimens exceeded the calculated Phase I, the energy dissipated per strength (mild) reinforcing steel in
values and tlle precast specimens per- cycle by the precast Zone 4 specimens conjunction with post-tensioning as a
formed as well as the monolithic was less than that of the monolithic means of improving the cyclic energy
specimens in most cases. Based on the specimens. Energy dissipation per dissipation characteristics of the pre-
results obtained thus far, the experi- cycle improved from the use of pre- cast concrete specimens. The premise
mental maximum moments for the stressing strands instead of post-ten- for tllis concept is that the mild steel
precast specimens post-tensioned with sioning bars with the post-tensioning will be used as an energy dissipater
strands were approximately 30 per- steel located closer to the beam cen- while the friction between the beam
cent greater than the predicted mo- troid, and fully bonded. The energy and the column caused by the post-

July-August 1993 91
tensioning force will be used to pro- greatly appreciated. Members of the to-Column Connections Subject to
vide the necessary shear resistance. steering committee are Robert E. En- Cyclic Loading," PCI JOURNAL,
To address the concern that the pre- glekirk, S. K. Ghosh, Daniel P. Jenny, V. 36, No.3, May-June 1991, pp. 56-67.
vious NIST tests did not have gravity (L. S.) Paul Johal and M. J. Nigel 5. Bhatt, Prabhakara, and Kirk, D. W.,
loads superimposed on the beams, Priestley. The authors would also like "Tests on an Improved Beam Column
to thank Suzanne D. Nakaki for pro- Connection for Precast Concrete,"
simulated gravity loads will be applied
ACI Journal, Proceedings, V. 82,
to the beams of the Phase IV speci- viding the design of the test specimens.
No. 6, November-December 1985,
mens. Other parameters that require pp. 834-843.
further study are the amount of post- 6. Bull, D. K., and Park, Robert, "Seismic
tensioning steel and mild reinforcing REFERENCES Resistance of Frames Incorporating
steel, the use of different concrete I. Uniform Building Code, International Precast Prestressed Concrete Beam
strengths and joint materials in the Conference of Building Officials, Shells," PCI JOURNAL, V. 31, No. 4,
construction joint, and the influence Whittier, CA, 1985. July-August 1986, pp. 54-93.
of slabs and transverse beams on con- 2. Cheok, G. S., and Lew, H. S., "Perfor- 7. French, Catherine, et al., "Connections
nection behavior. mance of /;-Scale Model Precast Con- Between Precast Elements - Failure
crete Beam-Column Connections Sub- Within Connection Region," ASCE
jected to Cyclic Inelastic Loads - Structural Journal, V. 114, No. 2,
ACKNOWLEDGMENT Report No. 1," NISTIR 4433, National February 1989, pp. 3171-3192.
The authors would like to express Institute of Standards and Technology, 8. Priestley, M. J. Nigel, and Tao, Jian
Gaithersburg, MD, October 1990. Ren, "Seismic Response of Precast Pre-
their gratitude to the individuals who
3. Cheok, G. S., and Lew, H. S., "Perfor- stressed Concrete Frames With Par-
contributed to this project. The assis-
mance of /;-Scale Model Precast Con- tially Debonded Tendons," PCI JOUR-
tance and support of the laboratory
crete Beam-Column Connections Sub- NAL, V. 38, No. 1, January-February
staff for the Building and Fire Re- jected to Cyclic Inelastic Loads - 1993, pp. 58-69.
search Laboratory, Structures Divi- Report No. 2," NISTIR 4589, National 9. ACI Committee 318, "Building Code
sion, especially that of Frank Rankin, Institute of Standards and Technology, Requirements for Reinforced Concrete
is gratefully acknowledged. Gaithersburg, MD, June 1991. and Commentary (ACI 318-89),"
In addition, the technical guidance 4. Cheok, Geraldine S., and Lew, H. S., American Concrete Institute, Detroit,
provided by the steering committee is "Performance of Precast Concrete Beam- MI, 1989.

92 PCI JOURNAL