Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
ABSTRACT: This paper presents the results of an experimental investigation into the seismic performance of
a composite moment resisting frame system that consists of reinforced concrete columns (with an embedded
steel shape for erection purposes) and composite steel beam-reinforced concrete slab sections. Quasi-static re-
versed cyclic loading was performed on six two-thirds-scale subassemblage specimens. Various detailing options
were evaluated to provide full moment connection in both orthogonal framing directions at the beam-column
joints. With appropriate joint detailing, specimens exhibited a desirable beam plastic hinge mechanism with
stable hysteretic response. Composite beam sections maintained near full composite behavior beyond code based
drift limits with good energy dissipation characteristics and were able to undergo large plastic rotation magni-
tudes. This framing system may represent a viable alternative for low-to-mid-rise structures in high seismic risk
zones.
Testing Procedure
The experimental test setup (Fig. 1) consists of two 490 kN
(110 kips) hydraulic actuators applied at the beam ends to FIG. 3. Joint Instrumentation
simulate lateral loading and a pneumatic cylinder to apply a
constant column axial load of 890 kN (200 kips) (near 10%
of the column’s nominal axial capacity). Simulated lateral
loading was applied by pushing and pulling on the framing
beams in either the continuous or discontinuous directions,
while restraining the top and bottom of the column (pinned
connections). Beam end actuators were operated in displace-
ment control to apply quasi-static, reversed cyclic loading to
the specimen using incremental displacement amplitudes (two
cycles at each displacement amplitude). Displacement ampli-
tudes are represented in terms of drift angle demands as de-
fined in Fig. 1(b). Specimens #1, #2, #3, and #5 were first
tested in the continuous beam direction (actuators applied to
continuous beams) until a reduction in specimen strength was
observed. At this testing stage, specimens were expected to
develop significant yielding and concrete cracking, as well as
deforming beyond the code drift limits for the life safety per-
formance level; however, specimens would not develop a large
degree of concrete crushing and would still possess significant
reserve strength (loading was halted at this level so that the
discontinuous framing direction could be evaluated). The spec-
imen was then rotated 90⬚ in the test setup, and the specimen
was loaded to failure in the discontinuous direction. Specimen
#4 was first tested in the discontinuous beam direction until a
reduction in specimen strength was observed and then loaded
to failure in the continuous beam direction. Specimen #6 was
loaded to failure in the continuous beam direction and then
loaded to failure in the discontinuous beam direction.
Specimen Behavior
Although the performance of all six specimens will be dis-
cussed, the measured response for specimens #2, #5, and #6
will only be shown. The results from these three specimens
highlight important response characteristics that were observed
in the testing program. The instrumentation used for the
component deformation measurements that follow is shown in
Fig. 3. FIG. 4. Specimen #2 Response
Fig. 4 shows the measured response for specimen #2 in both
the continuous and discontinuous loading directions. The
graph labeled ‘‘Total’’ displays the averaged force-drift angle ␦Support ⫹ ␦Column = (5 ⫹ 6)L /2 (3)
data measured at the points of load application at the ends of
␦Joint = [(2 ⫹ 3) ⫺ (5 ⫹ 6)]L /2 (4)
each beam (as recorded by the load cells and displacement
transducers of the hydraulic actuators). The other three graphs ␦Beam = ␦ Total ⫺ ␦Joint ⫺ (␦Support ⫹ ␦Column) (5)
show the response from each of the structural components of
the specimen (‘‘Column and Support,’’ ‘‘Joint,’’ and where 2, ␦3, ␦5, and 6 = measured rotations (rad) by clinom-
‘‘Beam’’), which make up the ‘‘Total’’ response. The force eters shown in Fig. 3; L = distance from point of load appli-
and drift angle values were calculated as follows: cation to joint center [1,372 mm (54 in.)]; and ␦ Total = dis-
placement measured at left beam end (measured by the
Force = (FLeft ⫺ FRight)/2 (1) actuator).
Drift Angle = 100␦/L (2) It should first be noted that through monitoring specimen
damage accumulation and rotations of the columns at the joint
where FLeft and FRight = left and right actuator forces, respec- interface and at the supports, columns of all specimens re-
tively; and ␦ = ␦Support ⫹ ␦Column, ␦Joint, ␦Beam, or ␦ Total, in which mained elastic throughout the testing procedure. The hysteretic
432 / JOURNAL OF STRUCTURAL ENGINEERING / APRIL 2000
behavior observed in the ‘‘Column and Support’’ graph was
due to the flexibility of the supporting frames (some small
movement of the supporting frame with respect to the labo-
ratory strong floor was visually observed) and rotational resis-
tance in the supports at the top and bottom of the column.
Specimens #1 and #2 performed in a similar manner. In the
continuous beam direction, the predominate inelastic behavior
occurred due to beam flexure, with minor inelastic joint re-
sponse. In the discontinuous beam direction, the beams re-
mained essentially elastic, and all damage occurred in the joint
region. Specimens #1 and #2 were originally designed to ex-
perience joint failure mechanisms when tested in the contin-
uous direction. Concrete cylinder tests revealed a higher than
expected compressive strength. After the joint strengths were
recalculated using the higher compressive strength value, it
was found that the joint panel shear and vertical bearing
strengths of specimen #1 were approximately equal to the
composite beam flexural strength, and the calculated joint
panel shear strength of specimen #2 was less than that of the
composite beam flexural strength. Both specimens, however,
experienced predominately beam yielding when tested in the
continuous direction. The ASCE Task Committee guidelines
(1994) that were used to calculate the joint strengths do not
take into account the effect of the RC slab, nor the transverse
beams. It may be that the presence of the slab and transverse
beams increased the effective joint width such that a larger
concrete strut mechanism developed to resist joint shear
forces.
Because the steel beams are not continuous through the RC
column in the discontinuous direction, elongation of the joint
may occur. Elongation is defined in this paper as when the
beams on either side of the joint rotate independently due to
flexibility in the connections (i.e., bolt slip) and is calculated
as follows:
Elongation = 2 ⫺ 3 (6) FIG. 6. Specimen #5 Response
Both specimens #1 and #2 experienced significant rotational
deformations of this type when tested in the discontinuous the FBPs away from the column face ensured that the plastic
direction. Fig. 5 displays the measured elongation during the hinges form in the composite beam-slab sections for both spec-
discontinuous direction test of specimen #2. imens in both loading directions. The length of the plastic
Specimens #3 and #4 included joint plates to ease con- hinge zone was observed to be about 1.5 times the depth of
structability and enhance joint performance. In the continuous the composite slab-girder section. Comparing the measured
direction, both specimens developed a beam hinging mecha- joint elongation response for the discontinuous direction of
nism with essentially elastic joint response. In the discontin- specimens #2 and #5 (Fig. 5) demonstrates the improved per-
uous direction, specimen #3 failed in the joint region due to formance of the octagonal joint detail with an additional top
bearing and elongation mechanisms, whereas for specimen #4, angle flange connection.
beam hinging occurred. The octagonal joint region of speci- The graphs labeled ‘‘Left’’ and ‘‘Right’’ (Figs. 7–9) display
men #4 outperformed the square joint region of specimen #3. the moment-rotation response for the left- and right-side com-
Specimens #5 and #6 utilized an octagonal joint detail sim- posite beam-slab sections, respectively. The moment was cal-
ilar to specimen #4. Specimen #6 had a discontinuous beam culated as the actuator force (FLeft or FRight) multiplied by the
on only one side of the joint—simulating an exterior connec- horizontal distance between the point of load application and
tion. Both specimens developed a beam hinging mechanism the FBP. The beam rotation was calculated as follows:
when tested in both the continuous and discontinuous direc-
tions (Fig. 6 shows the response from specimen #5). Locating Rotation = 1 ⫺ 2 (for left beam) (7)
Rotation = 3 ⫺ 4 (for right beam) (8)
The solid horizontal lines on the graphs indicate the cal-
culated plastic moment strengths of the composite beam-slab
sections (using measured material properties) for different ef-
fective slab widths bf. For positive bending, the concrete
within the effective flange width acts in compression; for neg-
ative bending, the concrete is cracked and the reinforcement
within the effective flange width acts in tension. An effective
slab width bf = 1,500 mm (60 in.) corresponds to an effective
slab width of one-quarter of the beam span length as specified
in the AISC design code (Manual 1994); bf = 380 mm (15 in.)
corresponds to an effective slab width equal to the width of
FIG. 5. Joint Elongations for Specimens #2 and #5 (Discontin- the column; and bf = 0 mm is representative of the steel beam
uous) alone. The continuous direction of specimen #6 had a slab only
JOURNAL OF STRUCTURAL ENGINEERING / APRIL 2000 / 433
FIG. 7. Beam Moment-Rotation for Specimen #2 FIG. 9. Beam Moment-Rotation for Specimen #6
Specimen Experimentalc
number
(#) Directiona P/Nb Uncorrected Correctedd Calculatedc,e
(1) (2) (3) (4) (5) (6)
2 C P 105 ⫻ 10⫺6 145 ⫻ 10⫺6 220 ⫻ 10⫺6
2 C N 67 ⫻ 10⫺6 81 ⫻ 10⫺6 95 ⫻ 10⫺6
2 D P 99 ⫻ 10⫺6 133 ⫻ 10⫺6 220 ⫻ 10⫺6
2 D N 62 ⫻ 10⫺6 74 ⫻ 10⫺6 95 ⫻ 10⫺6
5 C P 101 ⫻ 10⫺6 153 ⫻ 10⫺6 215 ⫻ 10⫺6
5 C N 87 ⫻ 10⫺6 123 ⫻ 10⫺6 95 ⫻ 10⫺6
5 D P 91 ⫻ 10⫺6 131 ⫻ 10⫺6 215 ⫻ 10⫺6
5 D N 84 ⫻ 10⫺6 117 ⫻ 10⫺6 95 ⫻ 10⫺6
6 C P 82 ⫻ 10⫺6 113 ⫻ 10⫺6 195 ⫻ 10⫺6
6 C N 60 ⫻ 10⫺6 75 ⫻ 10⫺6 84 ⫻ 10⫺6
6 D P 72 ⫻ 10⫺6 95 ⫻ 10⫺6 215 ⫻ 10⫺6
6 D N 66 ⫻ 10⫺6 85 ⫻ 10⫺6 95 ⫻ 10⫺6
a
Continuous (C) or discontinuous (D) loading direction.
b
Positive (P) or negative (N) bending.
c
Equivalent moment of inertia (m4).
FIG. 8. Beam Moment-Rotation for Specimen #5 d
Corrected for shear deformations.
e
Using measured material properties.
on one side of the beam, giving it a maximum effective width f
Cracked.
of 940 mm (37.5 in.).
Excluding the continuous direction of specimen #6 (exterior
configuration), it appears from strain gauges on the slab bars steel beam and slab reinforcement) was taken to be 200,000
and analytical strength calculations in both positive and neg- MPa (29 ⫻ 106 psi), and for the concrete was calculated as
ative bending that the effective slab width bf is equal to the 4,700 ( f ⬘)
c
1/2
[57,000( f c⬘)1/2] (‘‘Building’’ 1995), where f c⬘ is the
full effective slab width given in the AISC design specifica- concrete compressive strength taken from cylinder test data
tions. (Manual 1994). Where a slab was present on only one [measured in MPa (psi)]. The concrete slab was transformed
side of the steel beam (continuous direction of specimen #6), into an equivalent area of steel for calculating the moments of
the effective flange width was equal to column width for pos- inertia. The effect of shear deformations was accounted for in
itive bending. The discontinuous directions of specimens #5 obtaining the corrected experimental values (i.e., the uncor-
and #6 did not reach this strength value; however, this can be rected values assume all deformations are due to flexure). The
partially attributed to the damage (minor slab spalling) caused shear area used in the shear deformation calculations was
during the continuous direction test. The continuous direction taken as the area of the web of the steel beam [2.06 ⫻ 10⫺3
of specimen #2 achieved a maximum positive moment m2 (3.2 in.2)]. The relatively short length of the beams in the
strength that was slightly below the AISC calculated value. test specimens resulted in a significant amount of (calculated)
Recall, however, that this specimen was not tested to complete shear deformation—resulting in the noticeable differences be-
failure. If testing was continued to higher displacement levels, tween the uncorrected and corrected experimental moment of
this specimen may have achieved higher moment strength val- inertia values. The corrected experimental moment of inertia
ues. A strong case can, therefore, be made for the above state- values for positive bending were approximately equal to 70%
ment of the effective slab width being equal to the AISC value. of the calculated values (based on gross section properties with
Equivalent elastic moment of inertia values for the com- an effective flange width equal to the AISC value), with the
posite beam-slab sections were obtained from the measured exception of the discontinuous direction of specimen #6 (hav-
test data and compared with calculated values (Table 2). For ing a slab on only one side of the column reduced the force
the calculated values, Young’s modulus for the steel (both the transfer mechanism). The negative bending values, based on
434 / JOURNAL OF STRUCTURAL ENGINEERING / APRIL 2000
cracked concrete sections and slab reinforcement within the ACKNOWLEDGMENTS
full effective flange width, show a good correlation for all
This research forms part of the U.S.-Japan Cooperative Earthquake
specimens except #6. Research Program for Composite and Hybrid Structures. Funding was
provided by the National Science Foundation (Grant No. CMS-9632442
RESULTS through Dr. S. C. Liu), the Texas Engineering Experiment Station, and
Department of Civil Engineering at Texas A&M University, College Sta-
tion, Tex. These generous supports are gratefully acknowledged. The writ-
In specimens with enhanced joint detailing (specimens #5 ers are also grateful to the reviewers for their useful comments and sug-
and #6), maximum composite beam-slab moments were gestions.
reached at a specimen drift angle of approximately 2.5%, with
the RC slab showing little or no damage at this level. At spec-
APPENDIX I. REFERENCES
imen drift angle levels above 3.5%, the slab began to deteri-
orate, and the flanges and web of the steel beam began to ASCE Task Committee on Design Criteria for Composite Structures in
buckle—reducing the observed moment strength. The ASCE Steel and Concrete. (1994). ‘‘Guidelines for design of joints between
7-95 design load standard (‘‘minimum’’ 1995) specifies a max- steel beams and reinforced concrete columns.’’ J. Struct. Engrg., ASCE,
imum allowable story drift of 2.5% of the story height under 120(8), 2330–2357.
Bracci, J. M., Moore, W. P., and Bugeja, M. N. (1999). ‘‘Seismic design
seismic loading for the life safety performance level. The re- and constructability of RCS special moment frames.’’ J. Struct. Engrg.,
sults presented here demonstrate that these RCS subassembla- ASCE, 125(4), 385–392.
ges easily meet this requirement. The specimens tested had an ‘‘Building code requirements for structural concrete and commentary.’’
aspect ratio between inflection points of 1.0 (beam span to (1995). ACI 318, American Concrete Institute, Farmington Hills,
column span). In most real building structures, beam spans are Mich.
longer than column spans (aspect ratio between inflection Deierlein, G. G., Sheikh, T. M., Yura, J. A., and Jirsa, J. O. (1989).
‘‘Beam-column moment connections for composite frames: Part 2.’’ J.
points of 1.5–2.0). It can be shown that the imposed drift Struct. Engrg., ASCE, 115(11), 2877–2896.
angle demands for the specimen tests would correspond to Design guidelines for composite RCS joints. (1994). H. Kuramoto, S.
larger interstory drift percentages; thus, the results presented Mehanny, and G. Deierlein, translators, Composite RCS Structures
herein are conservative. Subcommittee, Architectural Institute of Japan, Tokyo.
For the purposes of composite beam design, the following Griffis, L. G. (1986). ‘‘Some design considerations for composite-frame
recommendations regarding the effective flange width are structures.’’ AISC/Engrg. J., Second Quarter, 59–64.
Kanno, R. (1993). ‘‘Strength, deformation, and seismic resistance of
made based on the experimental results. For positive and neg- joints between steel beams and reinforced concrete columns,’’ PhD
ative bending strength, the effective flange width is equal to dissertation, Cornell University, Ithaca, N.Y.
the value specified by AISC (Manual 1994), except for the Leon, R. T., Hajjar, J. F., and Gustafson, M. A. (1998). ‘‘Seismic response
case where there is a slab on only one side of the beam. In of composite moment-resisting connections. I: Performance.’’ J. Struct.
this case, the effective flange width is equal to the column Engrg., ASCE, 124(8), 868–876.
width for positive bending (no change for negative bending). Manual of steel construction—load and resistance factor design (LRFD).
(1994). American Institute of Steel Construction, Chicago.
In terms of positive bending stiffnesses, take 70% of the cal- ‘‘Minimum design loads for buildings and other structures.’’ (1995).
culated stiffness using the same effective flange widths as ASCE 7-95. ASCE, New York.
above. For negative bending, include the reinforcement within ‘‘NEHRP recommended provisions for seismic regulations for new build-
the AISC effective flange width if a slab is present on both ings and other structures: Part 1—Provisions and Part 2—Commen-
sides of the beam. When a slab is on only one side of the tary.’’ (1997). Publ. Nos. 302 and 303, Federal Emergency Manage-
beam, the contribution of the reinforcement to negative bend- ment Agency, Washington, D.C.
Nishiyama, I., Itadani, H., and Sugihiro, K. (1998). ‘‘Bi-directional seis-
ing stiffness should be ignored. For the purpose of elastic anal- mic response of reinforced concrete column and structural steel beam
ysis, rather than attempt to assign different stiffnesses to those subassemblages.’’ Proc., Struct. Engrg. World Conf., Elvesier Sciences,
regions of the beams in positive versus negative bending, it is New York.
recommended that a constant stiffness equal to the average of Sakaguchi, N., Tominaga, H., Murai, Y., Takase, Y., and Shuto, K. (1988).
the positive and negative bending stiffnesses be assigned to ‘‘Strength and ductility of steel beam-RC column joint.’’ Proc., 9th
the entire member. World Conf. on Earthquake Engrg., Vol. 4, 713–718.
Sheikh, T. M., Deierlein, G. G., Yura, J. A., and Jirsa, J. O. (1989).
The enlarged joint detail used in specimens #5 and #6 was ‘‘Beam-column moment connections for composite frames: Part 1.’’ J.
capable of transmitting increased shear forces, and the joint Struct. Engrg., ASCE, 115(11), 2858–2876.
plates were able to provide confinement to the joint region and Specifications and commentaries for composite steel floor deck. (1989).
allow the formation of a strut mechanism without the need for Steel Deck Institute, Canton, Ohio.
cumbersome joint reinforcement. Uchida, K., and Noguchi, H. (1998). ‘‘Nonlinear three-dimensional finite
element analysis of RCS frame.’’ Proc., 5th Joint Tech. Coordinating
Com. Meeting, U.S.-Japan Cooperative Earthquake Research Program
on Composite and Hybrid Structures, Tokyo.
CONCLUSIONS U.S.-Japan Planning Groups—Joint Planning Workshop. (1992). ‘‘Rec-
ommendations for U.S.-Japan cooperative research program—phase 5
This paper presented an alternative structural framing sys- composite and hybrid structures.’’ Tech. Rep. No. UMCEE 92-29, Uni-
versity of Michigan, Ann Arbor, Mich.
tem for low-to-mid-rise construction in high seismic risk areas. Viest, I. M., Colaco, J. P., Furlong, R. W., Griffis, L. G., Leon, R. T., and
This system consists of RC columns with an embedded steel Wyllie, L. A. (1997). Composite construction design for buildings.
shape for erection purposes and composite steel beam-RC slab ASCE/McGraw Hill, New York.
(on composite metal decking) beam sections.
The experimental testing presented demonstrates that with APPENDIX II. NOTATION
appropriate joint detailing, the developing mechanism during
seismic loading can be enforced in the composite beam- The following symbols are used in this paper:
slab sections away from the column face. These composite
beam sections are capable of developing plastic hinges that b = horizontal distance between actuators;
dissipate energy in a stable hysteretic fashion under large bf = effective flange width;
plastic rotations. The composite beams maintained near full FLeft, FRight = left and right actuator forces, respectively;
composite beam behavior beyond code based drift limits. Rec- f c⬘ = 28-day concrete compressive strength;
ommendations for the design of these beams were given. h = vertical distance between column pin supports;