2230

EFFECT OF DISTRIBUTED MASS ON EARTHQUAKE RESPONSE OF

REINFORCED CONCRETE FRAMES

Reza ALAGHEBANDIAN1, Shunsuke OTANI2 And Hitoshi SHIOHARA3

SUMMARY

This paper studies the effect of near field earthquakes on the response of reinforced concrete
buildings with a special emphasis on the influence of vertical component of ground motions on
nonlinear response of framed buildings with distributed lumped masses along girders. The vertical
ground motion excites the vertical vibration of floor slabs. In conventional analyses, floor mass is
assumed to concentrate at beam-column connections for simplicity. Thus, vertical vibration due to
distributed masses along slabs and girders is neglected. This paper compares the response of the
conventional lumped-mass model with the model incorporating distributed vertical masses along
girders subjected to horizontal and vertical components of near field earthquakes. A computer
program for the nonlinear earthquake response analysis of frames with distributed masses was
developed. The comparison revealed that the horizontal story displacement and story shear
responses were little affected by the vertical motion. On the other hand, the axial force of columns
and the vertical displacement of girders are significantly influenced when a building is subjected to
a large vertical ground motion. The effect of vertical motion on the axial force of columns was
more critical when the contribution of lateral seismic load was small such as in a low-rise building
or in the interior columns of an intermediate or high-rise building. It was also observed that the
vertical vibration could cause more fluctuation in column axial forces, the intensity of which was
found proportional to the intensity of vertical ground motion. Distributed mass under vertical
motion can significantly change the size of axial force fluctuation in columns, and consequently
the lumped mass model must be used conscientiously in estimation of column axial force. The
contribution of distributed mass and vertical motion to the axial force in a frame column needs
more studies.

INTRODUCTION

Design of a ductile frame dissipating seismic energy based on the concept of capacity design method [NZS-
1982, NZS-1995, and AIJ-1990] is well accepted by engineers. One of the basic requirements of the method is to
design a column with high degree of protection against yielding during severe earthquakes. The effect of vertical
motion on the design axial forces in columns is not considered explicitly in the NZS and AIJ guidelines. When a
building locates close to the epicenter and/or ruptured fault, it may simultaneously suffer non-attenuated
horizontal and vertical components of a ground motion. The vertical peak ground acceleration (PGA) of an
earthquake record may exceed the horizontal PGA. Moreover, in a near field region, the peak of vertical-to-
horizontal spectral ratio is even larger than the ratio of the PGA, especially at short period spans [Bozorgnia and
Niazi 1995]. On the other hand, vertical period of a building system/member is small and falls in a narrow range
of 0.05 to 0.26 second [Anderson and Bertero 1974, Kikuchi and Yoshimura 1984, Bozorgnia and Niazi 1998], a
period span which may correspond to a range of high vertical response spectra, particularly in near field regions.
The effect of vertical motion to the column forces may not be negligible. This paper is an attempt to study the
earthquake response characteristics of R/C frames located in a near-filed region.

1
The University of Tokyo, Tokyo, Japan, Email: alan@rcs.arch.t.u-tokyo.ac.jp
2
The University of Tokyo, Tokyo, Japan, Email: otani@rcs.arch.t.u-tokyo.ac.jp
3
The University of Tokyo, Tokyo, Japan, Email: shiohara@rcs.arch.t.u-tokyo.ac.jp
 ANALYTICAL MODEL AND EARTHQUAKE MOTIONS

A computer program for nonlinear earthquake response analysis of R/C frames with distributed mass along
girders was developed. Girders were divided into 10 segments along their axis and floor masses were lumped at
the internal nodes and at the centerline of columns. One component model [Giberson 1967] was used for girder
segments with symmetric moment distribution and for columns with assymetric moment distribution. The
constant gravity loads were applied gradually as vertical concentrated forces at the centerline of columns and at
the internal nodes along a girder within 100 loading steps prior to the real dynamic analysis. Only material
nonlinearity was included, and the geometrical nonlinearity was assumed to be negligible. A trilinear skeleton
curve was assumed for a moment-rotation relation at each member ends. The Sugano equation [Sugano 1970]
was used to estimate the post cracking stiffness. The post-yielding stiffness was assumed equal to 1% and 5% of
the initial stiffness for a girder-segment and a column, respectively. Axial stiffness was assumed to be linearly
elastic. Beam-to-column joints were assumed to be rigid with a finite length equal to column or beam widths.
The members were assumed to have infinite ductility, so that failure by attainment of the actual ultimate strength
or deformation capacity of a member should not be considered. A viscous damping proportional to the mass
matrix and instantaneous stiffness matrix was assumed. Damping was assumed equal to 0.05 in fundamental
horizontal and vertical vibration modes.
Twenty seconds of two near field strong 2500 2500
Northridge-Saticoy Kobe-JMA
ground motion records, the 1995, Kobe- NS 444.1 Gals
Acceleration response spectra, gal

NS 820.5 Gals
UD 785.0 Gals UD 333.2 Gals
2000 2000
JMA record and the 1994, Northridge-
Saticoy record were used. Absolute 1500 1500
acceleration response spectra of linearly
elastic system with 0.05 damping 1000 1000

subjected to the records are presented in

Figure 1. The peak vertical response 500 500

acceleration of Northridge-Saticoy record

0 0
is larger than that of the horizontal 0.01 0.1 1 10 0.01 0.1 1 10
Period, sec Period, sec
response, and falls at a period of 0.10
second; while the peak vertical response Figure 1: Acceleration response spectra of a linearly
acceleration of Kobe-JMA record is elastic system with 0.05 damping subjected to
smaller than the horizontal response, and near field records used in this study
occurs at a period of 0.25 second.

3. SINGLE-STORY ONE-BAY PROTOTYPE FRAMES

Five R/C plane frames with different span lengths of 4, 8, 12, 16, and 20 meters long were designed for a
combination of dead load of 36 kN/m, live load of 12 kN/m, and seismic coefficient Cb of 0.2 (Fig. 2).
Dimensions of frame members were determined for a reinforcement ratio of about 1%. Compressive strength of
concrete and yielding strength of steel were 30 and 400 MPa; and modulus of elasticity of concrete and steel
were 29.2 and 200 GPa, respectively. Member sections and natural period of frames are listed in Table 1. One
half of live load was considered to be effective as inertia mass in the horizontal direction. Total dynamic weight
of frames were 192, 384, 576, 768, and 960 kN for span lengths of 4, 8, 12, 16, and 20m, respectively. The
analysis was carried out for different cases of frames subjected to (a) vertical motion alone, (b) horizontal motion
alone, and (c) vertical and horizontal motions; with lumped mass model , and with distributed mass model.
b
G ird e r 50 50
Table 1: Frame parameters and characteristics
3 00 0

50

L
L b H D Natural periods (sec)
H

(m) (mm) (mm) (mm) Horizontal Vertical

50

C o nven tion al lu m pe d m a ss
(L M m o d el)
D 4 300 400 350 0.16 0.06
C o lu m n 50 50 8 400 550 500 0.13 0.11
3 00 0

50

12 500 700 650 0.11 0.14

U n it in m m
D

16 500 700 800 0.12 0.22

50

D istribu te d lu m p ed m a ss 20 500 700 900 0.08 0.36

(D M m o de l)

Figure 2: Analytical model and prototype structure

2 2230
 200
Figure 3 compares the axial force of column

Axial force of column, kN

LM (H + V)
in the frame with 4-meter bay subjected to 150 DM (H + V)
horizontal and vertical (H + V) components
100
of Northridge motion. The axial force is 17%
larger in compressive side for distributed 50 Northridge-Saticoy
mass model (DM) than that of lumped mass 4-m bay length
0
model (LM). The ratios of maximum 0 5 10 15 20
response axial forces to the initial design axial Time, sec
forces are illustrated in Figure 4. The
Figure 3: Axial force in column of frame with 4m bay
contribution of horizontal motion to the axial
force is small in long span frames. The axial

Axial force ratio in column, fraction of design load

Kobe-JMA NS 820 Gal UD 333 Gal
forces ratio to the design load are increased 2.0
LM (H + V) DM (H + V) DM (H)
about 50% and 75% under Kobe and 1.5
Northridge ground motions, respectively.
1.0
This indicates that the maximum axial forces
are proportional to the magnitude of vertical 0.5
motion. However, column axial forces in 0.0
some frames are much affected by dynamic Northridge-Saticoy NS 444 Gal UD 785 Gal
characteristics of the frame and vertical 2.0

motion than that of magnitude of the vertical 1.5

motion. The LM model may underestimate
1.0
the column axial forces under some
circumstances. The interaction response 0.5
diagrams at the top of frame column are 0.0
compared in Figure 5. The column seems to 4 8 12 16 20
be much affected by change in the bending Span length, m
moment than fluctuation in axial force in
Figure 4: Axial force ratio in columns as a fraction
interaction diagram of column (Fig. 5a). The
of initial design load
effect of distributed mass and vertical motion
can be observed in a close up view of the
Northridge-Saticoy NS 444 UD 785 Gals
response in Figures 5b and 5c. Axial force in 5000

column is significantly increased by the vertical Distributed Mass Model

DM (H)
4000
motion. Axial force and bending moment of DM (H + V) 4m

column are further increased by the effect of DM Yi

3000 e ld
model under vertical motion. The LM model, on su
rfa
Axial force, kN

ce
the other hand, can not take into account the effect 2000
of vertical floor vibration to the bending moment.
This can be concluded from upright shape of 1000 Designed load
Response
response interaction diagram of frame subjected to
vertical motion alone in Figure 5b. It can be seen 0

from Figure 5c that the column may yield by the

-1000
combination effects of vertical motion and -250 -200 -150 -100 -50 0 50 100 150 200 250
distributed mass. Moment, kN.m

Northridge-Saticoy NS 444 UD 785 Gals Northridge-Saticoy NS 444 UD 785 Gals

200 200
Lumped Mass Model Distributed Mass Model
LM (H) DM (H)
DM (V)
LM (V)
150 150 DM (H + V)
LM (H + V)
Axial force, kN

Axial force, kN

100 100

Designed load Designed load

50 50
rface
ce
Yeild surfa

Yeild su

4m 4m
0 0

-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150
Moment, kN.m Moment, kN.m

Figure 5: Effect of distributed mass on interaction response diagram at top of column

3 2230
 SIX-STORY TWO-BAY PROTOTYPE FRAME

A six-story two-bay R/C plane frame structure was designed to fulfill the requirements of the “Design Guideline
for Earthquake Resistant R/C Buildings Based on Ultimate Strength Concept” introduced by Architectural
Institute of Japan in 1990 [AIJ 1990]. The Prototype structure was designed for a seismic coefficient Cb of 0.25.
Geometry of sections were determined based on a linear analysis such that inter-story drift angle was less than
1/300 when frame was subjected to inverse triangular equivalent seismic design load. The location of planned
yield hinges, weight of structure, dead and live loads at floors, and cross section of columns and girders are
shown in Figure 6. Compressive strength of concrete and yielding strength of steel were 30 and 400 MPa; and
modulus of elasticity of concrete and steel were 29.2 and 200 GPa, respectively. Member reinforcements are
listed in Table 2. A nonlinear pushover analysis under triangular lateral seismic load was carried out to confirm
the location of planned hinges (Figure 6).Fundamental natural period of frame was 0.44 and 0.062 seconds for
horizontal mode and vertical column mode, respectively.

D L = 46 .2 k N /m 700
Table 2: Members reinforcements
G ra vity w e ig ht (kN )
L L =1 0 .8 k N /m (s e is m ic h oriz o n tal w e ig h t) 50 50 Bottom Top Slab
L L (s e is m ic)=4 .8 Story Member
6 8 4 .0
(6 1 2 .0 ) 50 steel steel steel
3500

D L = 33 .6 k N /m
700

L L =1 0 .8 k N /m 6 (hinge) Girder 2D19 2D22 4D13

L L (s e is m ic)=4 .8 5 3 2 .8 6 (no hinge) Girder 4D22 6D22 4D13
50

(4 6 0 .8 )
5 Girder 2D22 3D22 4D13
3500

C olu m n

5 3 2 .8 4 Girder 3D22 4D22 4D13

(4 6 0 .8 ) 1200 150
3 Girder 4D22 5D22 4D13
3500

50
21500 m m

2 Girder 4D22 6D22 4D13

700

5 3 2 .8
(4 6 0 .8 ) 1 Girder 5D22 5D25 4D13
50
3500

50 50
Hinge Column 12D22
U nit in m m
5 3 2 .8 350 No hinge Column 12D29
(4 6 0 .8 ) G ird er
3500

Inter-story drift angle, rad

5 3 2 .8 1/200 1/100 1/50
1600 0.55
(4 6 0 .8 )
1-floor
1400 Story shear 0.48
4000

2-floor

Base shear factor, Cb

1200 0.41
Story Shear, kN

3-floor
1000 4-floor 0.34
6000 6000
800 0.27
12000 5-floor
600 0.21

400 6-floor 0.14

Figure 6: Layout of the prototype structures,
200 0.07
location of planed hinges, and
0 0.00
results of pushover analysis 0.0 0.5 1.0 1.5 2.0
Inter-story drift angle, %

The analysis was carried out for different cases of the frame with distributed mass (DM) model under (a) vertical
motion alone (V), (b) horizontal motion alone (H), and (c) horizontal and vertical motions (H + V); and (d) the
frame with lumped mass (LM) model under vertical and horizontal motions. The frame was subjected to the
Kobe-JMA and Northridge-Saticoy records. Figure 7 compares the maximum lateral displacements and story
shears. Lateral displacements and story shears are little affected by the distributed mass and vertical motion.
DM (H)
6 6 DM (V) 6 6
LM (H + V)
DM (H + V)
5 5 5 5

4 4 4 4
Story

Story
Story

Story

3 3 3 3

2 2 2 2

1 1 1 1 Kobe-JMA
Northridge-Saticoy Kobe-JMA Northridge-saticoy
-20 -10 0 10 20 -30 -15 0 15 30 -1500 0 1500 -1500 0 1500
Lateral displacement, cm Lateral displacement, cm Shear force, kN Shear force, kN

Figure 7: Maximum lateral story displacements and story shears

4 2230
The maximum vertical displacements at mid-span of
frame girders are shown in (Figure 8). The Vertical DM (V) LM (H + V)
DM (H) DM (H + V)
displacements of girders are considerably affected by
distributed mass under vertical motion. The LM 6 6
model can take into account only the vertical
5 5
displacements concerning with the columns
elongation. The DM model, on the other hand, results 4 4
to larger vertical displacements by including the

Story
Story
vertical vibration of floor systems into the response. 3 3

It is observed that the vertical displacements are 2

2
magnified at floor level.
The maximum vertical displacements at the top of 1 1
Kobe-JMA, Girder mid-span
Northridge-Saticoy, Girder mid-span
interior and exterior columns are compared in Figure -0.5 0.0 0.5 -0.4 0.0 0.4
9. Interior columns are significantly affected by Vertical displacement, cm Vertical displacement, cm

vertical motion while the exterior columns are mainly

Figure 8: Maximum vertical displacements at
affected by horizontal motion. Response
the mid-span of frame girders
displacements due to vertical and horizontal motions
are comparable in exterior columns of the frame
subjected to Northridge record. This is because of extremely strong vertical component of this record. the LM
model reasonably estimates the vertical displacement at the interior columns of frame subjected to Kobe record
but overestimates the response when frame is subjected to Northridge record. Figure 10 compares the maximum
axial forces in the interior and exterior columns. Since the axial forces and the axial displacements in columns
were assumed to be linearly proportional, similar trends as to the vertical displacements are observed. The axial
forces in columns are significantly influenced by vertical motion, especially at interior columns. Distribution of
maximum axial forces is nearly linear over the height of the building. The LM model greatly overestimates the
axial force in interior column under Northridge record. The DM model may reduce or enlarge the maximum
axial forces in columns when comparing with the results of the LM model. Extremely strong vertical motion of
Northridge record results to tension in interior columns, while under Kobe record the interior columns are not
experienced tensile forces.

6 6 6 6

5 5 5 5

4 4 4 4
Story

Story

Story
Story

3 3 3 3

2 2 2 2

1 1 1 1
Northridge-Saticoy, Exterior column Kobe-JMA, Exterior column Northridge-Saticoy, Interior column Kobe-JMA, Interior column
-0.15 0.00 0.15 -0.15 0.00 0.15 -0.3 0.0 0.3 -0.06 0.00 0.06
Vertical displacement, cm Vertical displacement, cm Vertical displacement, cm Vertical displacement, cm

Figure 9: Maximum vertical displacements at the top of interior and exterior columns

6 6 6 6

5 5 5 5

4 4 4 4
Story

Story

Story

Story

3 3 3 3

2 2 2 2

1 1 1 1
Northridge-Saticoy, Exterior column Kobe-JMA, Exterior column Northridge-Saticoy, Interior column Kobe-JMA, Interior column
-1000 0 1000 2000 3000 -1000 0 1000 2000 3000 -2000 0 2000 4000 -1000 0 1000 2000
Axial force, kN Axial force, kN Axial force, kN Axial force, kN

Figure 10: Maximum axial force in the interior and exterior columns

5 2230
 Figure 11 compares the variations of base shear versus the top story lateral drift. It is observed that small
discrepancies exist under Northridge record but overall response of the frame is not sensitive to the vertical
motion and distributed mass. Lateral story displacements at first story of the frame are presented in Figure 12.
The lateral story displacement is very little affected by the vertical motion and the distributed mass.

1500 0.51 1500 0.51

Kobe-JMA
Northridge-Saticoy
1000 0.34 1000 DM (H) 0.34
DM (H)
LM (H + V)
LM (H + V)

Base shear factor, Cb

Base shear factor, Cb

DM (H + V)
DM (H + V)
500 0.17 500 0.17
Base shear, kN

Base shear, kN
0 0.00 0 0.00

-500 -0.17 -500 -0.17

-1000 -0.34 -1000 -0.34

-1500 -0.51 -1500 -0.51

-1.0 -0.5 0.0 0.5 1.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
Top story drift ratio, % Top story drift ratio, %

Figure 11: Overall earthquake response of prototype framed structure

6 6
DM (H) DM (H)
4 LM (H + V) 4 LM (H + V)
DM (H + V) DM (H + V)
Displacement, cm

Displacement, cm

2 2

0 0

-2 -2
Northridge-Saticoy Kobe-JMA
-4 First story lateral drift -4 First story lateral drift

-6 -6
0 5 10 15 20 0 5 10 15 20
Time, sec Time, sec

Figure 12: Lateral displacements at first story of the frame

Axial forces in exterior and interior columns at first story of the frame are compared in Figure 13. More
fluctuations in column axial forces are observed when the frame is subjected to vertical motion. The sizes of
fluctuations are larger under the Northrige record. This is because of stronger vertical ground motion of
Northridge record than the Kobe record. Interior columns are much affected by vertical motion than that of
exterior columns, which were mainly influenced by the overturning moment due to horizontal motion.
5000 2400
DM (H) 2200 DM (H)
4000
LM (H + V) LM (H + V)
DM (H +V) 2000 DM (H +V)
3000
Axial force, kN

1800
Axial force, kN

2000
1600
1000
1400
0
Northridge-Saticoy 1200 Kobe-JMA
-1000 First story interior column First story interior column
1000

-2000 800
0 5 10 15 20 0 5 10 15 20
Time, sec Time, sec

2500 2500
DM (H) DM (H)
2000 LM (H + V) 2000
LM (H + V)
DM (H + V) DM (H + V)
1500 1500
Axial force, kN
Axial force, kN

1000 1000

500 500

0 0
Northridge-Saticoy Kobe-JMA
-500 First story exterior column -500 First story exterior column

-1000 -1000
0 5 10 15 20 0 5 10 15 20
Time, sec Time, sec

Figure 13: Axial force in exterior and interior columns of first story

6 2230
Initial gravity load of interior and exterior columns were 1555 and 896-kN, respectively. The design axial load of
interior columns were calculated based on gravity load alone. The design lateral seismic load did not influence
the axial load of interior columns. Exterior columns were designed for combination of seismic and gravity loads.
The design axial load of exterior columns under unilateral seismic load was equal to –259 and +2019-kN. Table
3 listed the maximum fluctuation of axial forces in interior and exterior columns at the first story of the frame
under Northridge and Kobe strong motions. The numbers in parenthesis indicate the axial force fluctuation ratio
of response to the initial gravity axial load in columns. The value of axial force fluctuation ratio at the interior
column under Northridge strong motion is about 120% and 180% for DM and LM models, respectively. The LM
model significantly overestimates the axial force of an interior column. Under Kobe earthquake motion, the size
of axial force fluctuation ratio in interior column is about 40% for both DM and LM models. Exterior columns
are less affected by distributed mass and vertical motion. The contribution of vertical motion to the axial force
fluctuation ratio of exterior columns is about 10-20% under Kobe record, and about 40-60% under Northridge
record. The ratio is changed about 10% by using LM or DM models.
Figure 14 compares the interaction diagrams of interior and exterior columns. It can be seen that the maximum
axial forces can be occurred simultaneously with the maximum bending moment. This is clearly observed in the
interior columns under Kobe records. Consequently a column which is designed for the combination effects of
lateral seismic load and gravity load may yield by the effect of distributed mass under strong vertical motion.

Table 3:Maximum fluctuation in the axial force of interior and exterior columns
DM (H) LM (H + V) DM (H + V)
Member
Decrease Increase Decrease Increase Decrease Increase
Interior column 1480 1790 1033 2257 1000 2270
(Kobe-JMA) (5 %) (15 %) (33 %) (45 %) (35 %) (49 %)
Interior column 1546 1665 -1182 4522 -491 3414
(Northridge-Saticoy) (-1 %) (1 %) (176%) (-190 %) (125 %) (120 %)
Exterior column -572 2197 -690 2311 -782 2392
(Kobe-JMA) (164 %) (145 %) (177 %) (157 %) (187 %) (166 %)
Exterior column -241 1926 -571 2473 -687 2344
(Northridge-Saticoy) (126 %) (114 %) (163 %) (176 %) (176 %) (161 %)

4000 2500

3500 DM (H)
LM (H + V) 2000
3000
DM (H + V)
2500 1500
Axial force, kN

2000
Axial force, kN

1000
1500
500
1000

500 0
0 DM (H)
Northridge-Saticoy -500 Kobe-JMA LM (H + V)
-500 First story interior column DM (H + V)
First story interior column
-1000 -1000
-1500 -1000 -500 0 500 1000 1500 -1500 -1000 -500 0 500 1000 1500
Bending Moment, kN.m Bending Moment, kN.m

2500 2500

DM (H) DM (H)
ce

2000 2000
surfa
ce

LM (H + V) LM (H + V)
surfa

DM (H + V)
Yield

1500 DM (H + V) 1500
Yield

Axial force, kN
Axial force, kN

1000 1000

500 500

0 0

-500 -500 Kobe-JMA

Northridge-Saticoy
First story exterior column First story exterior column
-1000 -1000
-1500 -1000 -500 0 500 1000 1500 -1500 -1000 -500 0 500 1000 1500
Bending Moment, kN.m Bending Moment, kN.m

Figure 14: Interaction response diagrams of first story columns

7 2230
The moment-rotation relationships of the girder-end at the first story of frame are compared in Figure 15. It can
be seen that the effects of vertical motion and distributed mass on the moment-rotation relationship are very
small. This indicated that the variation in columns axial force due to vertical motion and distributed mass are not
affected by the change in shear force in the girders. Therefore the effect of girder shear force was negligible
when studying the effect of vertical motion and distributed mass onto axial force of columns.
800 800

600 600

400 400
Bending moment, kN.m

Bending moment, kN.m

200 200

0 0

-200 -200

-400 Northridge-Saticoy -400 First story girder end

First story girder end Kobe-JMA
-600 DM (H) -600 DM (H)
LM (H + V) LM (H + V)
-800 DM (H + V) -800
DM (H + V)
-1000 -1000
-0.002 -0.001 0.000 0.001 0.002 -0.003 -0.002 -0.001 0.000 0.001 0.002 0.003
Rotation, rad Rotation, rad

Figure 15: Moment-rotation hysteretic relationship at the exterior end of first story girder

5. CONCLUSIONS

The effect of distributed mass as well as vertical ground motions of two near field earthquakes to the nonlinear
dynamic response of five single-story one-bay R/C frames, and a six-story two-bay R/C frame were studied. It
was concluded that:
1. The lateral story displacements and story shears were slightly affected by the vertical motion and
distributed mass.
2. The vertical component of a ground motion significantly affected the axial forces in columns. This is
especially happened when the contribution of lateral seismic load on axial load is small such as in a low-
rise building or in the interior columns of an intermediate or a high-rise building.
3. Distributed mass under vertical motion can significantly alter the axial forces in frame columns. Thus
lumped mass model must be used carefully when estimating the design load of a non-yielding frame
column subjected to vertical motion. Further research is needed to estimate a magnification factor in axial
force fluctuation of column due to the effect distributed mass under vertical motion.
4. Vertical vibration caused more fluctuation in column axial forces. The size of fluctuation was proportional
to the intensity of vertical ground motion.

6. REFERENCES

1. AIJ (1990), “Design Guidelines for Earthquake Resistant Reinforced Concrete Buildings Based on
Ultimate Strength Concept.” Architectural Institute of Japan, Tokyo, in Japanese.
2. Anderson, J.C. and Bertero, V.V. (1974). “Effects of Gravity Loads and Vertical Ground Acceleration on
the Seismic Response of Multistory Frames,” Proceedings of 5th-WCEE, Vol.2, pp. 2914-2923, Rome.
3. Bozorgnia, Y., Niazi, M. and Campbell, K.W. (1995). “Characteristics of Free-field Vertical Ground
Motion during the Northrige Earthquake,” Journal of Earthquake Spectra, Volume 11, No. 4.
4. Bozorgnia, Y., Mahin, S.A. and Brady, A.G. (1998). “Vertical Response of Twelve Structures Recorded
during the Northridge Earthquake,” Earthquake Spectra, Volume 14, No. 3.
5. Giberson, M.F. (1967). “The Response of Nonlinear Multi-Story Structures Subjected to Earthquake
Excitation,” EERL Report, Earthquake Engineering Research Laboratory, California Institute of
Technology, Pasadena, California.
6. Kikuchi, K. and Yoshimura, K. (1984). “Effect of Vertical Component of Ground Motions and Axial
Deformation of Columns on Seismic Behavior of R/C Building Structures,” Proceeding of 8th WCEE, Vol.
4, San Francisco, USA, pp. 599-606.
7. NZS 3101 (1982, 1995). “Code of Practice for the Design of Concrete Structures,” Standards Association
of New Zealand.
Sugano, S. (1970). “Experimental Study on Restoring Force Characteristics of Reinforced Concrete Members,”
Thesis submitted to fulfill the requirements of Ph.D. degree, University of Tokyo, in Japanese

8 2230