• ......

.
tv)

N •
Oo
.:::t-
QUASI-STATIC TESTS OF A SCALED MODEL BUILDING

by Seung-Joon Lee 1 and Le-Wu Lu2 , Member, ASCE

ABSTRACT: The testing of a 0. 305 scale model of the

six-story prototype steel building under a program of

controlled cyclic displacements is described. A "wiffle

tree" system was used to produce a lateral load pattern

closely resembling the pattern specified by the UBC code.

The testing was carried out with three lateral load

resisting schemes: dual system with concentric braces,

dual system with eccentric braces and moment-resisting

space frame. The overall performance of the three

systems is discussed and the test results are compared

with theoretical predictions.

INTRODUCTION

A 0.305 scale model of the six-story steel prototype building structure

was constructed and tested in a quasi-static manner at Fritz Engineering

Laboratory, Lehigh University, as part of the overall investigation (Foutch,

et al. 1987, Roeder, et al. 1987). In this study, emphasis is on the

overall performance of the structure to strong earthquake- induced lateral

forces and the hysteretic characteristics of the structure. Three types of

lateral force resisting schemes were investigated:

Phase I Dual system with concentrical braces in Frame B (CBF).

1Asst. Prof. of Architecture, Ajou Univ. Suweon, Korea, formerly Res.

Assoc., Dept. of Civ. Engrg., Lehigh University, Bethlehem, PA 18015
2Prof. of Civ. Engrg., Lehigh University, Bethlehem, PA 18015

1
..

Phase II Dual system with eccentrical braces in Frame B (EBF).

Phase III - Moment-resisting frame system (MRF) with Frame B unbraced.

It is recognized that the earthquake-resistant design of buildings must

satisfy two basic conditions: (1) The building must maintain its

serviceability during moderate earthquakes for which sufficient lateral

strength and stiffness are required, and (2) It must not collapse during a

severe earthquake. The second condition imposes the require~ents of

deformability and the associated energy dissipation capacity. Moment

resisting frame (MRF) is known to exhibit good deformability and energy

dissipation characteristics, but it lacks the strength and stiffness at

large lateral deflections. Concentrically braced frame (CBF) has sufficient

stiffness and is widely used for tall buildings. However, under a major

earthquake, buckling of the braces can cause severe loss of energy

dissipation capacity and deterioration of strength and stiffness.

Eccentrically braced frame (EBF) is known to perform well under both

moderate and severe earthquakes, because it is a stiff and strong structure

and has an excellent energy dissipation capacity.

Many types of structural components and subassemblages have been tested

to investigate their load-deformation behavior. Most of the frame tests

were conducted on planar structures. Very few tests have been performed on

three dimensional complete structures. The seismic behavior of dual systems

is not well understood because of the complexity of the interaction between

the braced and unbraced frames. The overall behavior of the dual system,

after failure of the primary lateral force-resisting system, has not been

fully investigated, especially in the large deformation range.

2
,.
' .

MODEL STRUCTURE

The scale factor of 0. 305 for the model structure was selected by

considering the available laboratory floor space and the scale factors used

in other tests of the cooperative program. The dimensions of the model are

shown in Fig. 1. A dimensional analysis was employed to derive the

similitude laws between the physical dimensions of the quantities involved

(Lee, 1987). The most difficult task to fulfill is the requirement of

higher density material for the model in a manner that does not change the

strength and stiffness characteristics. This requirement may be satisfied

by placing additional distributed weights to achieve the desired dynamic

characteristics. Because the model tests were conducted statically and '

because the dead load effect is expected to be small, the extra weights were

not used.

The structural members were fabricated to be exact replicas of those of

the prototype. Model members were manufactured using long strips which were

cut from 4 x 8 ft.(l.2M x 2.4M) steel plates or sheets. For theW-shapes,

two flange strips and one web strip were welded together using the TIG

welding process. For the tubular braces, two strips bent at 90 degree were

welded to form a square or rectangular section. The average yield stresses

of the steel sheets (gage #7 to #12) and steel plates for W shapes were

32.36 ksi (223.3 MPa) and 43.97 ksi (303.4 MPa), respectively, and the

average yield stress of the brace material was 63.62 ksi (370.0 MPa).

Light weight concrete with the density of 115 pcf (1840 kg/m 3 ) was used

for the prototype floor slab. For the model structure, solite aggregate

with a maximum size of 0.25 in. (6.4 mm) and normal sand were used. A total

of 96 6 x 12 in. (150 x 300 mm) test cylinders were prepared during casting

of the slabs. The average weight was 102 pcf (1630 kg/m 3 ), the average

3
compressive strength at 28 days was 2. 87 ksi (19. 8 MPa), and the average

modulus of elasticity of some selected cylinders was 2200 ksi (15190 MPa).

A galvanized steel deck was used to simulate the prototype H. H.

Robertson's QL-99 deck. The thickness was 0.033 in. (0.84 mm) and the rib

height, average rib width and distance between ribs are 1. 0, 2. 0, and 4. 0

in. (25, 50 and 100 mm), respectively. The deck dimensions were 9% larger

than those required for perfect modeling. Since the model deck did not have

embossments, screws were projected into the concrete to develop the

necessary bonding. Shear studs with a diameter of 0.25 in. (6.4 mm) and a

height of 1. 56 in. (39. 6 mm) were installed on all the girders. A welded

wire mesh with a diameter 0.043 in. (1.09 mm) and a spacing of 0.5 in. (12.7

mm) and a yield stress of 60 ksi (414. 0 MPa) was used for the slab

reinforcement.

All other details of the prototype were simulated as closely as

possible. The column-to-column connection used full penetration butt welds

in the flanges and webs. For the connections where column sections of

different depths met, a rectangular plate of 0.5 in. (12.7 mm) thickness was

inserted to ensure continuity. In the girder-to-column moment connections,

fillet welds were provided on both sides of the flanges and the web, and, in

the shear connections between girder and floor beam, only the web was

welded. The details of the brace-to-girder connections are discussed in the

relevant sections that follow.

EXPERIMENTAL BEHAVIOR AND RESULTS

In each phase of the study a flexibility test was first performed and

followed by cyclic tests using displacement control.

Loading Program: The flexibility test was conducted by applying lateral

4
..

loads, one at a time, to the individual floors and measuring the resulting

deflections. In the cyclic tests, the desired lateral loading pattern was

the pattern prescribed by the Uniform Building Code (Uniform Building Code,

1976), on the bases of which the prototype structure was designed (Askar, at

el., 1983). The desired pattern was achieved with a two jack loading system

and a specially constructed "wiffle tree". Figure 2 shows the pattern of

the lateral loads applied to the test structure and the UBC pattern. At

each floor level, the load was divided into two equal loads applied to the

slabs through the triangular frames (marked s2 in Fig. 1). The horizontal

displacement at the roof of Frame B was used as the control in the cyclic

tests.

Instrumentation: The model structure was instrumented to obtain the

lateral load vs. deflection response and to determine the behavior of the

critical structural components. Because of the limitation of the available

data acquisition systems, it was not possible to instrument all the

structural members and to measure their deflections. The lateral

deflections at the floor levels, and elongations of selected braces were

measured by linear variable differential transformers (LVDT). Strain gages

were attached on the columns of Frame B to determine the moments, axial

forces and shear forces. The bracing force was measured by a pair of strain

gages mounted at a selected section. Clip gages were placed on some girder-

to-column joints to measure the distortion of the panel zone. The applied

forces were monitored by two load cells placed between the hydraulic jacks

and the wiffle tree system.

Phase .! Testing: After a couple cycles of trial tests had been performed

to check the instruments and the data acquisition systems, flexibility

matrix of the structure was established by the flexibility test. The matrix

5
is used later to compute the natural frequencies which are listed in line 1

to Table 1. For the cyclic tests, a total of 19 cycles of controlled

displacements were applied and the test data were recorded at a total of 755

load steps (LS). Figure 3a shows the roof displacement vs. total base shear

curves of the cyclic tests. Figures 3b and 3c show, respectively, the

individual floor deflections and the total base shear at all the load steps.

The response was essentially linear, until the applied load reached about

the design shear of 24.8 kips (110.3 kN) for the model structure. The first

yielding was observed in the girder web of the brace-to-girder joint at the

second floor at LS 72. The corresponding base shear was 39.6 kips (178 kN).

More yielding was observed at the girder webs of the brace-to-girder joints

at the third and fourth floors, as the load was increased in the subsequent

cycles. The details of the joints were believed to cause high shears to

develop in the girders. The secant stiffness of the model at LS 123 was

about 85 percent of the initial tangent stiffness and a small hysteresis of

the load-deflection curves was observed.

At LS 157, the brace. at the south side of the 1st story started to

buckle. The total base shear was 49.84 kips (221.7 kN). In the following

cycle of loading, this brace buckled severely, and the other brace, which

was under tension, fractured at its upper connection. The fracture was

repaired and both braces were strengthened by welding narrow steel strips to

the four sides, which resulted in a 56 percent increase in area and 5

percent increase in radius of gyration. The test was then resumed. Figure

4d shows the cyclic behavior of a 3-story brace. At LS 361, the south side

brace of the 5th story buckled out of plane, and at LS 365 the same brace of

the 4th story buckled in plane. In the subsequent cycle, the opposite

braces of the 4th (LS 388) and the 5th stories (LS 391) buckled in and out

6
of plane, respectively. When the roof displacement reached 1.65 in. (41.9

mm), the panel zones of the strong axis oriented girder-to-column joints was

observed, but no visible yielding was in the girders and columns.

The structure was then subjected to cyclic roof displacements of 1.9,

2.1 and 2.2 in. (48.3, 53.3 and 55.9 mm). At LS 450, the lower connection

of the brace at the south side of the 4th story fractured. The structure

was unloaded. After repairing the cracks, the structure was reloaded to the

roof displacement of 1.9 in. (48.3 mm). During the loading cycle, the south

side brace of the 2nd story buckled in a bi-axial direction at LS 501, and

both braces of the 3rd story buckled in plane at LS 561 and LS 592. The

maximum base shear was 62.42 kips (277.6 kN) at the roof displacement of 2.1

in. (53.3 mm). The braces at the 4th and 5th stories started to crack at

the corners of the severely kinked sections and the cracks propagated across

the sides (Fig. 4). During the later cycles of loadings, the load dropped

with increasing roof displacement. The brace at the 4th story finally

ruptured and the load reduced to 80 percent of the maximum base shear at the

roof displacement of 2.4 in. (61.0 mm). At the roof displacement of 2.6 in.

( 66. 0 mm) , the base shear was only 77 percent of the maximum base shear

reached and the test was terminated. Figure 5 shows the damage pattern of

the model structure after test. The slabs on several floors cracked at

sections adjacent to columns in the direction transverse to the loading and

at areas above the brace-to-girder joints.

Phase II Testing: After the completion of the phase I test, the concentric

braces were removed and the eccentric braces were installed in the south bay

of Frame B for the Phase II test. Figure 6a shows the details of the brace-

to-girder connection. The gap between brace and girder flange was smaller

than that of the prototype, which was believed to be responsible for the

7
observed out-of-plane buckling of the joint. Two strain gages were attached

on each brace to measure the axial force.

The same test procedure as that of Phase I was adopted for Phase II.

Line 2 of Table 1 lists the natural frequencies of the structure computed

from the results of the flexibility test. For the cyclic tests, a total of

15 cycles of loadings with 0.2 in. (5.1 mm) increment were applied and data

were recorded at a total of 572 load steps (LS). Figure 6b gives the total

base shear vs. the roof displacement relationships, and Figs. 6c and 6d show

the individual floor deflection and the total base shears at all the load

steps. As expected, the shear link at the 2nd floor exhibited yielding

first. At the roof displacement of 0.6 in. (15.2 mm), the web of the link

showed flake-off of the whitewash. At the roof displacement of 0. 8 in.

(20. 3 mm), the shear yielding spreaded over the whole web, and the shear

links at the 3rd and 4th floors also yielded. At the roof displacement of

1.0 in. (25.4 mm), all the links up to the 6th floor exhibited yielding.

None of the columns and girders exhibited yielding until the roof

displacement reached 1. 8 in. ( 45. 7 mm) . Unlike the Phase I test, the

strength of the structure continued to increase as the applied roof

displacement increased. Some 1st story columns near the footing showed

yielding at the subsequent cycles of loadings.

When the roof deflection reached 2.2 in. (55.9 mm), the severely

distorted link at the 2nd floor started to crack at the edge of the welds of

the intermediate stiffeners. The links at the 3rd and 4th floors cracked at

the same locations in the next cycle of loading. As the applied

displacement increased, the cracks grew and rupture finally occurred at 3.0

in. roof displacement. The test was then terminated. Figure 7 shows the

fractured web of the 3rd floor link. Yielding in many girder-to-column

8
joint panel zones of the moment frames were also observed.

Phase III Testing: Before beginning the test, all the damaged shear links

were repaired and a thorough visual inspection of the structure was made.

The flexibility test gives the natural frequencies listed in line 3 of Table

1. A total of 25 cycles of displacements with a 0.2 in. (5.1 mm) increment

were applied. Figure Sa shows the cyclic relationship between total base

shear and the roof displacement. Figures 8b and 8c show the individual

floor deflections and the total base shears at all the load steps (LS). The

structure had many weak panel zones. Panel zones of some of the interior

beam-to-column joints exhibited yielding at the roof displacements of 1.2

and 1.4 in. (30.5 and 35.6 mm) and yielding of additional panel zones in

Frames A and C was observed in the following cycles. Figure 8d shows the

cyclic moment vs. angular distortion relationship of a panel zone. Some

columns at the column-to-footing joints showed yielding at 1.8 in. roof

displacement. The test was terminated at the roof displacement of 5 in.,

which was the maximum capacity of the LVDT used in controlling the test.

DISCUSSION

Overall Response: Figure 9 shows the skeleton curves of the roof

displacement vs. total base shear of the three phases of testing. The

elastic limit strength is defined as the strength when the first yielding or

buckling was observed in a structural member. Table 2 lists the elastic

limit strengths and the ultimate strengths of the three structures.

The Phase I test structure (CBF) has large elastic limit strength, but

it has the smallest reserve strength beyond the elastic limit. Buckling of

a brace did not cause sudden drop of the load. The strength and stiffness,

however, deteriorated rapidly and the structure showed a descending load-

9
deflection relationship. The rupture of the braces limited the overall

ductility and the energy absorption capacity. The shares of the applied

loads resisted by the moment-resisting frames increased significantly after

the braces had failed. The current design practice requires that the CBF

system be designed with the largest base shear of the three systems.

The Phase II test structure (EBF) exhibited bilinear type of load-

deflection relationship with stable hysteretic behavior. The elastic

stiffness was about 10 percent higher than that of the Phase I structure as

estimated by the natural frequencies. The 2nd slope of the load-deflection

curve initiated when a shear link started to yield. Because of the high

reserve strength of the links beyond initial yielding, this structure shows

the largest reserve beyond the elastic limit strength. The resistance of

the structure, however, decreased rather significantly when rupture occurred

in a severely distorted link. The behavior of the shear links controlled

the overall strength and ductility of the structure. The overall ductility

was estimated to be about 4.0. The dual system with EBF performed well.

The Phase III test structure (MRF) is a very flexible structure. The

elastic stiffness was only 20 to 22 percent of those of the two dual

·systems. Most of the inelastic deformation was in the panel zones of the

girder-to-column joints, which were . somewhat underdesigned (Lee and Lu,

1988). The structure exhibited gradual stiffness degradation after first

yielding, but the hysteresis loops were very stable. A ductility of about

5.0 was reached at the termination of the test.

Story Shear vs. Storv Drift: Table 3 summarizes the maximum story shear

force and story drift of the individual stories reached during the three

phases of testing.

In the Phase I testing, the 1st story brace buckled early and was

10
strengthened. Then the 4th story underwent the most inelastic deformation.

After buckling of its braces. Fig. lOa shows the story shear vs. story

drift curves of this story. The damage was concentrated in this relatively

weak story, in relation to the amount of story shear applied.

In the Phase II testing, the inelastic deformation developed in the

shear links according to the expected sequence, that is, starting from the

lower stories then moving upward. This behavior is considered highly

desirable for earthquake resistant design in that the damage was distributed

uniformly among the stories and the stable hysteretic characteristics of

shear links were fully utilized. The cyclic behavior of the 3rd story of

the structure is shown in Fig. lOb.

Figure lOc shows the 2nd story shear vs. drift relationships of the

Phase III test structure. It is a very flexible structure and exhibited

stable hysteretic behavior.

Behavior of Braces: The hysteretic relationship between the axial force

and axial deformation of a brace of the Phase I test structure is shown in

Fig. 4d. In the elastic range and before brace buckling, the axial forces

in the pair of braces in any story were approximately equal, one being in

tension and the other in compression. The compression brace eventually

buckled either in an in-plane or out-of-plane mode. The resistance to axial

compression of the buckled number decreased rather rapidly and an increased

share of the story shear was then carried by the tension brace. The buckled

brace was never fully straightened under tension when the direction of the

applied lateral displacements was reversed. After a few cycles of buckling

and partial straightening, the thin-walled tubular braces started to tear at

the corners, which eventually led to fracture. In this test structure,

brace fracture occurred in the 4th and 5th stories.

11
 Table 4 lists the bracing forces measured at the time of buckling and

the calculated buckling force using Formula (1.5-1) of the American

Institute of Steel Construction (AISC) Specification (Specification for the

design, fabrication and erection of structural steel for buildings 1978)

without the safety factor and with an effective length factor K-0.65. This

comparison shows that strong rotational restraints existed at the ends of

the braces. The braces of the Phase II test structure were so designed that

they would not buckle even after the shear links had reached strain

hardening.

Behavior of Link Beams : The cyclic shear force vs. shear distortion

relationship of the 3rd floor shear link is plotted in Fig. 11. The link

showed generally stable hysteretic behavior and large energy dissipation

capacity. It's stiffness decreased only gradually after initial shear

yielding. The maximum shear resistance reached at a 3% angular distortion

and remained essentially constant until extensive cracks developed in the

link web. Its distortion capacity was about 10 percent. The concrete slab

above the link cracked rather severely and there was some loss of composite

action between the steel girder and the slab at large distortions. The

details of the brace-to-girder connections, which were an improvement of the

details used in the prototype structure, resulted in stable behavior of the

braces and the gusset plates.

Behavior of Panel Zones: Yielding of panel zones was observed in all the

three phases of testing. In the two dual systems, the panel zones started

to yield after the braced frame underwent significant inelastic deformation,

which occurred due to buckling of the braces or yielding of the shear links.

In the Phase III test, the inelastic deformation of the panel zones was very

dominant and represented an important energy absorption mechanism. It has

12
been suggested (Krawinkler, et al. 1971 and Lee, 1987) that limited

inelastic deformation of the panel zones may be regarded as a desirable

feature in seismic-resistant building structures, because it reduces the

ductility demand of other members in the structure. Analytical hysteretic

models for steel and composite joints with panel zone deformation have

recently been developed (Lee, 1987, Lu, et al. 1988).

Distribution of Story Shear Force: Figure 10 has been plotted to show the

change of the distribution of the story shear to the braces (if any), the

braced frame and the moment frames of the three test structures. In the

Phase I test, initially, the braced frame and braces took a slightly larger

portion of the story shear than predicted. The buckling of the braces

caused rapid transfer of the shear to the moment frames. In the 4th story,

the two moment frames resisted about 60 percent of the story shear after the

pair of braces failed by fracture. In the elastic range of the Phase II

test, the shear taken by the braces and the braced frame was 84 percent

which is about the same as in the Phase I test. However, unlike the Phase I

structure, yielding of the shear links caused only a small increase of the

portion of the story shear resisted by the moment frames. The distribution

of the story shear to the three moment frames remained essentially constant

throughout the Phase III test. This indicates that it is possible to treat

the structure as three individual planar frames for purpose of analysis.

EXPERIMENTAL VS. ANALYTICAL

Static incremental inelastic analyses were performed on the test

structures and the results were compared with the experimental results. The

DRAIN-20 program (Kannan and Powel, 1973), after modified for statically

applied lateral loading, was used in these analyses. Also, two subroutines,

13
one for the composite beam element and the other for the composite beam-to-

column joint element with panel zone deformation, were incorporated into the

program. The details of these elements are presented elsewhere (Lee, 1987).

For simplicity, only the skeleton curves of the roof displacement vs. base

shear relationships of the test structures are presented here. The results

of the detailed studies will be given in a forthcoming report.

Figure 12a shows the comparison of the results of the Phase I test.

Two analytical curves are shown. Analysis 1 was performed before the test

and analysis 2 after the test. Analysis 2 included the shear beam elements

at the brace-to-girder joints which were subjected to high shear and was

yielded during the tests and elastic spring elements at the bottoms of the

columns of the braced frame to simulate the flexible foundation. The

stiffnesses of the springs were determined from the readings of the dial

gages mounted on the column bases. The load-deflection relationship from

Analysis 2 shows good correlation with the experimental results.

In the analytical study of the Phase II structure, the most critical

element is the shear links. The test result exhibited a trilinear behavior

of the shear links. In the DRAIN-20 program, only bilinear model is

available. The analytical results using the bilinear model resulted in

discrepancies between the test and the analysis, especially in the large

deflection range, as shown in Figure 12b. Improved modelling of composite

shear link behavior is needed at this time.

Figure 12c shows the experimental and analytical skeleton curves of the

Phase III test. The close correlation indicates the adequacy of the models

for composite beam and composite beam-to-column joint developed in this

study.

14
SUMMARY

A 0. 305 scale model building structure was tested cyclically using a

quasi-static test method. Three types of lateral load resisting systems

were studied. An inverted triangular lateral load pattern according to the

UBC code was used in the test. The results are summarized as follows:

1. The overall hysteretic behavior of the three types of structures

was as expected. The design details of the connections, however,

played an important role in affecting the performance of the

structures. The brace-to-girder connection details of the CBF

caused early yielding in the beam. The details of the brace-to-

girder connection of the EBF adopted were successful in preventing

out-of-plane buckling of the braces and the gusset plates.

2. Buckling of braces in the CBF caused a marked reduction of the

stiffness of the Phase I test structure, but its load-carrying

capacity did not reduce too rapidly until one of the braces i,n a

story ruptured. The dual system with CBF showed less ductility and

small reserve strength beyond the elastic limit.

3. Hysteretic behavior of the shear links was very stable. The dual

system with EBF showed superior performance with regard to

ductility and energy absorption capacity. The ductility of the

system was limited by fracture of the links. The maximum

distortion of link beam reached was about 10 percent.

4. The moment-resisting space frame was very flexible. Extensive

yielding of the panel zones of the girder-to-column joints occurred

with any fracture. The overall hysteresis behavior was stable and

the ductility was large.

15
 5. Quasi-static test method using an inverted triangular lateral load

distribution caused damages concentrated in a particular story in

the Phase I test. For the Phases II and III tests, however, the

damages were more distributed.

6. The analytical load-deflection curves presented show good agreement

with the test results. There is a need to develop improved

behavior models for shear links.

ACKNOWLEDGMENT

This research described in this paper was supported by the National

Science Foundation through Grant No. CEE-8207712 and was part of the "U.S.-

Japan Cooperative Earthquake Research Program Utilizing Large-Size Testing

Facilities," initiated by Joseph Penzien and Hajime Umemura. During the

course of investigation, the program was coordinated by Robert D. Hanson and

Makoto Watabe. The valuable support and advice provided by Drs. Michael P.

Gaus, S. C. Liu and John B. Scalzi of the NSF are gratefully acknowledged.

The findings and opinions expressed are those of the authors and do not

necessarily represent the views of the sponsor.

APPENDIX ! ~ REFERENCES

Askar, G., Lee, S. J., and Lu, L. W. (1983). "Design studies of the six-

story steel test building". Report No. 467 .3, Fritz Engrg. Lab., Lehigh

Univ., Bethlehem, PA.

Foutch, D. A., Goel, S. C., and Roeder, C. W. (1987), "Seismic testing of a

full-scale steel building- part I". J. Struct. Engrg., ASCE, 113(11), 2111

- 2129.

16
Kannan, A. E., and Powell, G. H. (1973). "DRAIN-2D: A general purpose

computer program for dynamic analysis of inelastic plane structures".

Report No. UCB/EERC 73-6, Earthquake Engineering Research Center, Univ. of

California, Berkeley.

Krawinkler, H., Bertero, V. V., and Popov, E. P. (1971). "Inelastic

behavior of steel beam-to-column subassemblages". Report No. UCB/EERC 71-7,

Earthquake Engineering Research Center, Univ. of California, Berkeley.

Lee, S. J. (1987). "Seismic behavior of steel building structures with

composite slabs". Thesis presented to Lehigh Univ. at Bethlehem, PA in

partial fulfillment of the requirements for the degree of Doctor of

Philosophy.

Lee, S. J. and Lu, L. W. (1988). "Cyclic tests of full-scale composite

joint subassemblages". Report No. 467.7, Fritz Engrg. Lab., Lehigh Univ.,

Bethlehem, PA.

Lu, L. W., Wang, S. J., and Lee, S. J. (1988). "Cyclic behavior of steel

and composite joints with panel zone deformation". Paper presented at the

Ninth World Conference on Earthquake Engineering, Tokyo, Japan, (To appear

in the conference proceedings).

Roeder, C. W., Foutch, D. A., and Goel, S. C. (1987). "Seismic testing of a

full-scale steel building - part II". J. Struct. Engrg., ASCE, 113(11),

2130 - 2145.

Specification for the design, fabrication and erection of structural

steel for buildings (1978). American Institute of Steel Construction,

Chicago, ILL.

Uniform Building Code (1976). Int. Conference of Building Officials,

Whittier, CA.

17
 TABLE ! NATURAL FREQUENCIES OF MODEL STRUCTURE

Frequencies, Hz
Test
Structure First Mode Second Mode Third Mode
(1) (2) (3) (4)

Phase I 2.835 8.422 14.13

Phase II 2.985 9.030 17.17

Phase III 1. 323 3. 779 6.436

TABLE .2, STRENGTH OF MODEL STRUCTURE

Test Elastic Limit Ultimate

Structure Strength, kips Strength, kips
(1) (2) (3)

Phase I 49.84 62.42

Phase II 38.36 73.68

Phase III 15.08 34.64

Phase I Phase II Phase III

Story Drift Story Drift Story Drift

Story Shear, Index Shear, Index Shear, Index
kips kips kips
(1) (2) (3) (4) (5) (6) (7)

6th 15.54 0.0045 18.35 0.0028 8.63 0.0074

5th 31.21 0.0193 36.84 0.0089 17.32 0.0207

4th 43.19 0.0203 50.99 0.0113 23.97 0.0245

3rd 52.37 0.0127 61.82 0.0158 29.06 0.0264

2nd 58.80 0.0083 69.41 0.0167 32.63 0.0235

1st 62.42 0.0065 73.68 0.0137 34.64 0.0148

18
 BUCKLING STRENGTH OF BRACES

Buckling Strength, kips

North South AISC

Story Brace Brace (K-0.65)
(1) (2) (3) (4)

5th 18.27 17.23 14.65

4th 19.63 23.41 21.70

3rd 26.71 25.36 27.80

2nd -- 34.07* 27.80

1st -- 25.36 21.82

* Brace connection was reinforced

19
 90.0

:t;====o~=====rrc=,=====o;--===·: c,
I II I

~--2·----~~-----1't-~
0
0
0'1
I
1G 5
I

·--------~
I
I
I I
I I
II

11
II
""ool
11
II

. -----
o. #,-
J::• _..,.
I
Direction
of Loading
I I Gt
·:8}+--
I II
==========~=-=
I C• 1•ac:1 II Cs
...,..··===
Gt
I
I C. ~
I ii :: l
0 ~---"'----~~----,..--!
:o, •. 1:.. •.....
II ..........
0
0'1
~I --- -----~. -----.11:-~
f .. ,

. II
# ..

CAr-t-- ::':::=G~:=:dl;~:~=:-- I c,
0 UNIT:inches
1..0
FLOOR PLAN
FRAME 8
Phase I

r1
I I I
~ ~ ~
FRAME B FRAME 8
Phase II Phase III

Fig. 1. Floor Plan and Elevations of Model Frame B

 0.249 R

.. 0.251 6
.241

+---0.192 5

.--0.147 4

~0.103
3

•o.ose 2
UBC-. 1.00 1.00 ._TEST

Fig. 2. Lateral Load Distribution

 BASE SHEAR DEFLECTION (IN)
75.0 ].

(b)
(a) 50.0 2.

I.

-3.0 -2.0 2.0 3 0 70 800

DEFLECTION LS
-I.

-50.0 -2.

-75.0 -3. FLOOR DEFLECTION HISTORY

FORCE BASE SHEAR

45.0 75.0

(c)
50.0

2!1.0

-.80 . .?0 .40 .60 80 DO . 100 800

DEFORMATION LS
-2!1. 0

-30.0 3RD STORY -50.0

v IJ IJ

-45.0 -75.0
LOADING HISTORY

Fig. 3. Cyclic Behavior of Phase I Test Structure

(a) Roof Deflection vs. Base Shear (b) Floor
Deflection History (c) Loading History
(d) Axial Deformation of Brace
 Fig . 4 . Fracture of Brace

FRAME AS.C FRAME 8

BEAM A~O COLUMN BRACE
,.:::..-Fiang• • in-plan• buckling
'<:::1'--.w • b • out-of- plan• buckling
e bi -aKial buckling
X fractur•

Fig . 5 . Damage of Phase I Test Structure

 ..
DEFLECTION liN!

(a) {c)
..
..

-I .
-
LS

....
-I.

FLOOR DEFLECTION HISTORY

BASE SHEARIKIPS) BASE SHEARIKIPSI

71. 71.

(b) ..
(d)
J I I
...
-c. -1.

-IS.
r..
-
LS

....
\
Fig. 6.
-71. -71.

LOADING HISTORY
Phase II Test (a) Detail of Shear Link Connection
(b) Cyclic Roof Deflection vs. Base Shear
'
(c) Floor Deflection History (d) Loading History
Fig . 7 . Fracture of Shear Link
 ..
BASE SHEAAIKIPSl DEFLECTION (IN)

.. (b)

eoo 100 1000

s
·2 .

...
.... ...
FLOOR DEFLECTION HISTORY

...
I\40MENT IK IPS- IN) BASE SHEAR (KIPS)
....
(d) ... (c)

......
·I! .
- - ...
....
.... .... LOADING HISTORY

Fig. 8. Cyclic Behavior of Phase III Test Structure

(a) Roof Deflection vs. Base Shear (b) Floor
Deflection History (c) Loading History
(d) Distortion of Panel Zone
 BASE SHEAR (KIPS)
75.

50.

2!1. ,.-.. -----.u --- -----

...... ......,-
,.,.""'
-5. ·-4. -3. -2. -1. .......... 2. 3. 4. 5.
....... ROOF DEFLECTION(IN)
...... ......
---- ----- -- --
_,;
-2!1.

-so.

·-·- /
-75.

Fig. 9. Skeleton Curve of Roof Deflection vs. Base Shear

 STORY SHEAR RATIO
7S 0

(a) ~0 0

1.00 - I~ . '50 /5 1.00 aoo

STORY DRIFT LS

4TH STORY

-7~.0
lOt-----------------
-
...._
llo----4>
STORY SHEAR I MAX. S. S.
BRACE 0 F./ S. S.
BRACE IS. S.
STORY SHEAR(KIPS) RATIO
7S.

(b) !50.

.5

1.00 -.~ .!50 . 7S 1.00

STORY DRIFT (IN)

.5
--- 540
LS

3RO STORY
1.0 -------------------------
- STORY SHEAR I MAX. S. S.
-7S.
...._ BRACED F./ S. S.
llo----4> BRACE I S. S.
STORY SHEAR(KIPSJ RATIO
45.

t.O - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -.::-.;-,_,-.-..--
(c)
JO.

LOAD TO NORTH

-.1~ .50 .IS S.OO 100 200 JOO 500 700 BOO 900 1000
STORY DRIFT (INJ LS

.~

-JO. 2ND STORY

1.0 ---------------------------------
......... STORY SHEAR /MAX. S. S.
~ FRAME B/S. S.

Fig. 10. Story Shear vs. Story Drift and Distribution

of Story Shear
 SHEAR (KIPS)
30.

.12
N (Rad.)

-30.

Fig. 11. Distortion of Shear Link

 IASI SMURC-ipal
10

ll.t.SI SMEAR C-ipal

__,~:~·-- 60

I ,.. ''
I

ANAI..1--...... i /
so -..z '"I"" 1/
LOADING TD NOQT M

40

JO
Phase I Phase II

U'IRIIENTAL
ANALYTICAL
- IXNRIIENTAL
-ANALYTICAL

lO 1.0 z.o JO
11001' DII'L£CTIDNCinl

BASE SHEAR!kips)
40

Phase III
30

20 LOADING TO SOUTH

EXPERI~ENTAL

ANALYTICAL

1.0 2.0 3.0 4.0 5.0 6.0

ROOF DEFLECTION(in)

Fig. 12. Experimental vs. Analytical Results

 ,
T'

Steel, moment frame, braced frame, beam, column, bracing, connection,

buckling, stiffness, composite slab, earthquake resistance.

SUMMARY

The results of tests of a scale model of a prototype six-story steel

building structure are presented. The test program included three lateral

load resisting schemes: dual system with concentrical braces, dual system

with eccentrical braces and pure moment frames.