You are on page 1of 11

13

th
World Conference on Earthquake Engineering
Vancouver, B.C., Canada
August 1-6, 2004
Paper No. 2224


NON-LINEAR NUMERICAL ANALYSES TO IMPROVE THE SEISMIC
DESIGN METHOD FOR SOFT FIRST STORY RC BUILDING


Kheir-Eddine RAMDANE
1
, Koichi KUSUNOKI
1
, Masaomi TESHIGAWARA
1
and Hiroto KATO
1



SUMMARY

Soft first story buildings are one of the most vulnerable structural types during severe earthquakes. Such
structures, notably RC buildings are required in overpopulated areas.

Kobe earthquake, which is considered to be one of the most devastating and costly natural disaster in recent
history, bearing in mind the number of buildings destroyed, the number of people killed and injured, and the
damage extended to a wide range of structural types. It was found that many buildings that were constructed
with open retail space or parking on the first floor collapsed.

The collapse of the soft first story structure was attributed to inadequate transverse reinforcement in terms of
its amount and detailing, more flexible and/or weaker story and to relatively smaller amount of shear walls to
allow for the access to the open space than the ones above. Hence, the Building Standard Law Enforcement
Order of Japan, revised after the Kobe earthquake, adopted stricter guidelines for the construction of soft first
story buildings. That was a tentative approach, but no reliable methodology for the design of this type of
structure exists so far.

The aim of this study is to evaluate the seismic capacity and conduct non-linear numerical analyses of typical
RC buildings of 6-, 10- and 14 stories with soft first story, investigate the response characteristics and
improve the seismic design method of this type of structures. An expression is derived based on the energy
constant law for the strength increasing factor, and compared with results from the numerical analyses to
verify the adequacy of this expression.

INTRODUCTION

Ground shaking produced by an earthquake will search for any structural vulnerability. These vulnerabilities
are usually created by sharp changes in stiffness, strength and/or ductility, lack of shear walls in the first
story, and the effects of these vulnerabilities are accentuated by poor distribution of masses. As a result, one
of the most vulnerable structural types is the multistory commercial and residential buildings built of
reinforced concrete with little or no lateral resistance of the first story. These buildings are called Piloti or
soft first story buildings.

1
Building Research Institute, 1 Tachihara, Tsukuba, Ibaraki, Japan 305-0802. Email: eddine_r@yahoo.fr

Generally, soft stories have less stiffness than the upper stories. Together with a lack of ductile capacity in the
reinforced concrete columns, beams, and joints, these soft stories can cause many brittle failures during an
earthquake ground motion. The 1978 Miyagi-ken Oki Earthquake caused brittle and severe damage of the
columns of the first floors to RC buildings, which had shear walls except in the first floors. This type of
damage was more noticed in the 1995 Kobe Earthquake. Based on this fact, a design procedure which does
not allow column side-sway mechanism (soft first story collapse) was recommended in the Japanese Seismic
Design Guidelines.

On the other hand, soft first story collapse might be permissible even in Piloti buildings if the maximum
response deformation angle of the first story can be kept to the level of the maximum response deformation
angle supposed in beam sideway mechanism (total collapse) buildings. And since soft first story buildings
have fewer and/or no shear walls to absorb the energy generated by an earthquake than those total collapse
buildings, thus in order to hold the maximum response deformation angle of soft first story buildings to the
same level as that of total collapse buildings, and according to the energy constant law, the horizontal load-
carrying capacity required in the soft first story buildings must be increased in comparison with total collapse
buildings.

In this paper, a procedure to determine the failure modes of a building with a soft first story from the failure
modes of columns and walls at the soft story is demonstrated. This will be used to develop a design
procedure to secure the safety of those buildings allowing the soft first story collapse mechanism. Then a
strength increasing factor,
p
, which represents the capacity demand for the soft story in order to get the
same total collapse of the system, is determined based on the energy constant law so that the ductility factor
for buildings with a soft first story collapse,
'
, is equal to the ductility factor for buildings with total
collapse mechanism,

.

The buildings considered in this study are designed according to the specification prior to the 1995 Japanese
code. The values of the base shear coefficient C
b
divided by the vibration factor R
t
are 0.43, 0.46, and 0.43,
for the considered RC buildings of 6-, 10- and 14 stories with soft first story, respectively. Then the base
shear is varied with respect to the original value and the stiffness of the first story is assumed as constant.


OUTLINE OF THE SEISMIC DESIGN CODE IN JAPAN

Significant developments occurred in the Japanese Seismic Design Code since the restoration of Meiji period
in 1968. Following the 1923 Kanto Earthquake, the Urban Building Law, which concerned only six major
cities in Japan, required the use of a seismic coefficient of 0.1 in conjunction with a materials safety factor of
3 on ultimate stress. These requirements were revised from time to time following devastating earthquake
disasters.

The first Building Standard Law, applicable to all buildings in Japan, was revealed in May 1950. The
purpose of the law is to safeguard the life, health, and property by providing minimum standards concerning
the site, structure, equipment, and use of building. The Law requires that a building owner must submit,
before construction work, an application to building officials to confirm that the building, plan, site, structure
and equipment satisfy the provisions of the law. This requirement made the code prescriptive because
building officials must be able to judge the conformity. The structural design was based on the allowable
stress design.

The 1978 Miyagi-ken Oki Earthquake which caused serious damage, led to accelerated implementation of the
revised Building Standard Law in 1981. The Building Standard Law included a seismic coefficient that
varies with structural vibration period, and introduced a two-level design procedure. The first level design
follows the traditional allowable stress design approach, with steel allowable stress equal to the yield stress,
and concrete allowable stress equal to 2/3 of the specified compressive strength. The second level design is a
direct and explicit evaluation of strength and ductility, and maybe regarded as a check of whether these are
sufficient for severe ground motion.

In the first level design, the seismic (elastic response) coefficient C
i
at each floor level i is determined as:

0
C A R Z C
i t i
= Eq.(1)

where, Z: seismic zone factor, R
t
: vibration characteristic factor, A
i
: factor representing vertical distribution of
the seismic story shear coefficient, C
o
: basic base shear coefficient (0.2 for allowable stress design and 1.0 for
the examination of story shear resisting capacity).

Further, in the second level design, the story shear resisting capacity must be greater than the required
horizontal load-carrying capacity
un
Q at the soft story, which is given by:

=
i i s e S un
W C F F D Q Eq.(2)

where, D
s
: structural characteristic factor (0.30 0.55), representing the ductility of hinging members of the
story, F
e
: coefficient representing the eccentricity in a story, F
s
: representing the distribution of stiffness
along the height, C
i
: story shear coefficient,
i
W : total dead and live loads above story i. However,
buildings not more than 31 m and satisfying certain conditions are not required to examine the story shear
resisting capacity. Other structures of heights between 31 m and 60 m, must be checked by both design
procedures. Structures over 60 m in height are subject to special approval.

When the first story is ductile, D
s
become small, and so is the required horizontal load-carrying capacity. In
this case the whole building dissipates the energy uniformly. However, a soft story building needs to
dissipate almost all the earthquake energy at the first floor. If Eq.(2) is applied to the soft first story building,
the latter shows a large deformation. Hence, it is proposed that the maximum ductility or maximum
deformation angle of the soft first story building must be kept to the same level as for the total collapse
building. Therefore, considering the fact that the damage concentrates in the soft first story, for the required
horizontal load-carrying capacity of the first story, the larger of
p
and F
s
, as shown in Eq.(3), should be
used instead of F
s
in Eq.(2).

max
) , (
s p
F Eq.(3)

where
p
is the strength increasing factor to limit the deformation of the soft first story.

FAILURE MECHANISM OF SOFT STORY BUILDING

Inspections of earthquake damage as well as the results of analytical studies have shown that structural
systems with a soft story can lead to serious problems during severe earthquake ground shaking.

For instance, Fig.1 and 2 illustrate such damages. Fig. 1 shows the failure mechanism of soft story building.
These are: a) Bending (tensile yielding of reinforcing bar), b) collapse of first story (yield in column), and c)
collapse of first story (shear failure of column). As for a soft story with walls, two types of failure mechanism
are observed in a frame with a wall: a) bending (bending yield at wall bottom), and b) shear collapse of first
story (shear failure). The failure mechanism of the frame with wall is predominant and therefore controls the
failure mechanism of the whole system (building).


(a) (b) (c)

Fig. 1: Failure types of soft story building



(a) (b)

Fig. 2: Failure types of soft story with walls
Based on the above observed types of failure, the following section illustrates the Model which is
considered in this study.

MODEL FOR THE STRUCTURE

The model considered in this study is shown in Fig. 3. Continuous shear walls from the second to the top
story are modeled as equivalent one column as shown in Fig. 3(a), and each mass represent the weight of
each floor. Columns in the first story are modeled as two columns. The total weight of the first story is
distributed into three masses and each mass of the first story is equivalent mass. Beams in the first story
are modeled as a rigid element.

Rigid beam Rigid beam
Wall
Wall
Wall
Wall
Wall
Column Column

Rigid- plastic spring
for flexural behavior
Rigid- plastic spring
for flexural behavior
Linear spring
for shear
Linear spring
for axial behavior
Elastic behavior Elastic behavior

(a) Model for structure (b) Model for member
Fig. 3: Model for structure and member

For the first story, the shear and the axial springs are assumed to be elastic as shown in Fig. 3(b). As for the
flexural spring, the Takeda Model is applied. For the second story and above, the shear, the axial and the
flexural springs are assumed to be linear. Moreover, the applied damping is proportional to the instant
stiffness of 5%.
INPUT MOTIONS

The NS component of the record at the El Centro, Hachinohe, and Kobe were used as the input earthquake
motions in addition to four artificial waves (wg60, wg61, wg62 and wg63). Each record is normalized so
that the maximum velocity is 500 mm/sec
2
. The Peak Ground Acceleration (PGA) of each record is shown
in Table 1 and the response acceleration spectra for each of the seven input motions is shown in Fig. 4.

Table 1: PGA of each record

Record PGA (cm/sec2)
El Centro NS 448.12
Hachinohe NS 277.91
Kobe NS 444.05
Wg60 484.30
Wg61 467.33
Wg62 522.92
Wg63 510.04


0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0


R
e
s
p
o
n
s
e

A
c
c
e
l
e
r
a
t
i
o
n

(
m
/
s
e
c
2
)
Period (sec)
El C entro N S
Hachinohe NS
Kobe N S
Artificial W ave 1
Artificial W ave 2
Artificial W ave 3
Artificial W ave 4


Fig. 4: Response acceleration spectrum for each input motion


METHOD FOR DERIVING THE FACTOR
p


The energy constant law as shown in fig. 5 was considered when investigating the responses of soft first
story and total collapse mechanism buildings. The assumed magnitude of external forces is based on an
elastic response of 1.0G. For strengths lower than 1.0G, the building yields. As a result, and assuming
that the energy constant law is satisfied, the shear strength coefficient D
s
and ductility can be expressed
by Eq.(4) given as:

1 2
1

s
D
Eq.(4)
Furthermore, fig. 5 shows the relationship between the strength and the deformation for a one-mass
system. Assuming that the constant energy law is satisfied, it can be shown for instance that the Work W
(1)

= W
(0.3)
. Also, in cases where all stories yield in the same manner and the restoring force characteristics of
each story constitute a perfect elasto-plastic system, the ductility of a one-mass system is identical to the
ductility of each story of multi-mass system.

0 1 2 3 4 5 6
0.0
0.5
1.0
One- mass system
W
(0.3)
W
(1)
(0.3)
(0.5)
(1)

1y
S
t
r
e
n
g
t
h


(
G
)
Displacement ratio ( /
1y
)


Fig. 5: Energy constant law

As for the factor
p
, it is determined so that the ductility factor
'
for column sidesway building is equal
to the ductility factor

for beam sidesway building. It is assumed that the required internal work for both
types of buildings is the same and the distribution of external load and displacement is an inverted triangle.
The weight and the height of each floor are the same for both types of building. The structural
characteristic index is
'
S
D
for the soft first story building and
S
D
for the total collapse building as shown
in Fig. 6. Furthermore, the hysteresis for the i-th floor is assumed as an elasto-plastic model, and each
floor has the same stiffness for both buildings. Hence, the ratio of
'
S
D
to
S
D
to obtain

=
'
can be
expressed by Eq. (5);

S
S
p
D
D
'
=
Eq.(5)

By assuming n is the number of stories in the total collapse system; m and h are the weight and height of
each story, respectively; and are equal for both types of collapses; and
y
h
is the yield displacement
(story drift); and also by considering an inverted triangular distribution of the external force-deformation as
shown in Fig. 7, the internal work for the total collapse system, W
in
, is expressed by Eq. 6 given below:

( )
( ) ( ) 1 2 1
4
1
1
2
1
1
+ =

+ =

=


y S
n
i
y y S in
h D m n n
h Ds m i h D m i W
Eq. (6)

where, D
s
: is the demand story shear coefficient defined in the 2001 Japanese Seismic Design Regulation
(C
b
for first story), and : is the maximum ductility factor to assume a total collapse system. The
displacement of each floor is described in Fig.8.

Ds


Ds


(a) Total collapse system

(b) Soft first story collapse

Fig. 6: Total collapse system and soft first story collapse

As for the soft first story collapse system, almost the whole deformation is concentrated at the first floor,
hence the distribution shape shown in Fig. 9 is considered. It is assumed that the weight and height are
uniform in all stories, and are the same as in the total collapse system. Thus, the internal work done by the
soft story is given by Eq. (7):

( )
' ' ' '
1 '
2
1
y S y S
h D m n h D m n +
Eq. (7)

where,
'
y
h : is the story drift at yielding for soft first story collapse system, D
s

: is calculated as D
s

p
,
and

: is the maximum ductility factor assuming a soft first story collapse.




Fig. 7: External force-deformation
distribution for total collapse system.

imDs
h
y
h
y


Fig. 8: Restoring force characteristics of
respective stories.

The upper stories of the soft first story building are assumed elastic. This elastic work with respect to the
total work changes depending on the capacity and stiffness of each floor. Here the distribution shape is
assumed to be an inverted triangle as shown in Fig. 10. As can be seen from the figure, the upper stories
perform no work for =0.0 and if =1.0, upper stories can dissipate the same amount of energy as the first
story, which is elastic. Hence, the elastic work of the i-th story is expressed by Eq. 8 given as:
' '
) 1 (
2
1
y S
h D m i n +
Eq. (8)

where : is the coefficient representing the rate of elastic work load contribution above the first story of the
soft first story building.



Fig. 9: External force-deformation distribution
of soft first story collapse system.

=0 =1 0< <1


Fig. 10: Distribution shape

Accordingly, the total internal work for a building with a soft story, W
in
, is expressed as:

( )
( )
)
`

\
|
+ =
+ + =

2
1
' 1
4
1
1 '
2
1
2
1
'
' '
' ' ' '
1
' '


n h D m n
h D m n h D m n h D m i W
y S
y S y S
n
i
y S in
Eq. (9)

And as shown in Fig. 11 below, it assumed that the rigidity of each story in the total collapse system is the
same as for the soft first story collapse systems, thus the following relationship given by Eq.(10) can be
obtained:

S
S
y
y
D
D
'
'
=

Eq. (10)
imDs
h
y
imDs
h
y

imDs
h
y



Fig 11: The relation between
y
and
y


Since the internal work for the total collapse and soft first story collapse systems is equal and using Eqs.
(6), (9), and (10), Eq.(11) for
p

can be expressed as:



( ) ( )
( ) ( ) 1 ' 2 2 1
1 2 1
'
+
+
= =

n
n
D
D
S
S
p
Eq. (11)

Fig. 12 shows
p
for buildings up to 20 stories when 0 . 5 ' = = . The value becomes larger as the total
height increases. As can be seen from the figure, a value
p
for a 10 story building with a soft story is
equal to 1.92.3. This means that the required lateral load resistance of the soft first story collapse
system is about 1.9 to 2.3 times as large as that of the total collapse system when ductility is to be limited
to 5.0.

0 5 10 15 20
1.0
1.5
2.0
2.5
3.0
3.5
=1.0
=0.0

0.0 0.1
0.2 0.3 0.4
0.5 0.6 0.7
0.8 0.9 1.0


Fig 12: The relationship between
p

and the number of stories




RESULTS AND DISCUSSION

The obtained results of the D
s
- relationship for 14-, 10- and 6-story buildings are shown in figs. 13, 14
and 15, respectively. The scatter values present the results from the non-linear numerical analyses for the
seven input motions. Plots presenting the variation of the coefficient from 0 to 1.0 are drawn in the
figures. Also the constant energy and constant displacement plots are included.

0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6
0
2
4
6
8
10
12
14
16
18
20
Constant
displacement
C onstant
energy
w g60
w g61
w g62
w g63
El C entro
H achinohe
Kobe
=1.0
=0.0
14F

Ds


Fig. 13: D
s
-Ductility relationship for 14-story
0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6
0
2
4
6
8
10
12
14
16
18
20
Constant
displacement
Constant
energy
w g60
w g61
w g62
w g63
El C entro
H achinohe
Kobe
=1.0
=0.0
10F

Ds


Fig. 14: D
s
-Ductility relationship for 10-story
0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6
0
2
4
6
8
10
12
14
16
18
Constant
displacement
Constant
energy
w g60
w g61
w g62
w g63
El C entro
H achinohe
Kobe
=1.0
=0.0
6F

Ds


Fig. 15: D
s
-Ductility relationship for 6-story


For 14- and 10-story buildings, for ductility 0 . 6 = , a value of D
s
equal or greater than 0.6 is obtained
from the non-linear numerical analyses as shown in figs. 13 and 14, respectively. For a value of 0.6 or
greater, the response becomes small enough, however for a value D
s
of 0.4, the response is much greater
and gets close to a value of
p

of about 1.0, requiring, as a result, a high ductility. In the design of new


structural elements (columns) of soft first story building, it is recommended therefore, that a value of D
s
of
0.6 or greater to be applied to avoid a soft first story collapse.

On the other hand, for a 6-story building, and for ductility 0 . 6 = , a value of D
s
equal or greater than 0.55
can be applied in the design of new structural elements (columns) of soft first story building to avoid a soft
first story collapse. The result from the non-linear numerical analyses presenting the relationship (D
s
)
is shown in fig. 15.

Furthermore, fig. 16 below summarizes the D
s
-Deformation angle relationship for 6-, 10-, and 14-story. It
can be seen clearly that the deformation angle decreases as D
s
increases.

Finally, fig. 17 shows the relationship between the number of stories and
p
x D
s
for 6-, 10-, and 14-story.
As shown earlier, the value of the required lateral load factor
p
x D
s
is 0.55 for 6-story building and
increases to a value of 0.60 for 10-story building or greater.

0.4 0.5 0.6 0.7 0.8
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16


D
e
f
o
r
m
a
t
i
o
n

a
n
g
l
e

(
R
a
d
)
Ds
6F
10F
14F

Fig 16: D
s
-Deformation angle relationship for
6-, 10-, and 14-story.
0 5 10 15
0,25
0,30
0,35
0,40
0,45
0,50
0,55
0,60
0,65
0,70
0,75
0,80
Theoretical line
Sim plified line

p

x

D
s
Number of Stories

Fig 17: Number of stories-
p
x D
s
relationship
for 6-, 10-, and 14-story
CONCLUSIONS

A new seismic design procedure that could secure the structural safety of soft first story buildings during
severe earthquake motions by allowing column sideway mechanism at the first story has been proposed.

A relationship between the required design strength of the soft first story and seismic response of Piloti
buildings is clearly obtained by an analytical study.

From the non-linear numerical analyses and for ductility 0 . 6 = , a required lateral load factor
p
x D
s

equal to 0.55 is obtained for 6-story building, and increases to a value of 0.60 for 10-story building or
greater.


ACKNOWLEDGEMENT

Dr. Kheir-eddine Ramdane expresses his appreciation to the Japan Society for the Promotion of Science
(JSPS) for the two-year post-doctoral fellowship at the Building Research Institute, Tsukuba, Japan.
Special thankfulness is also extended to his colleagues at the BRI; Dr. Kusunoki, Dr. Teshigawara, Mr.
Kato, Dr. Fukuyama and Dr. Fukuta for their uncountable assistance and support.


REFERENCES

1. 2001 Technical Guidelines for Building Structure (in Japanese). The Building Centre of Japan.
2. Architectural Institute of Japan (1991). Standard for Structural Calculation of Reinforced Concrete
Structures. 1991.4.
3. Housner, G.W. (1952). Intensity of Earthquake Ground Shaking near the Causative Fault. The
Third World Conference on Earthquake Engineering, New Zealand, 94-115.
4. Otani, S. et al., Analysis of the Full Scale Seven Story Reinforced Concrete Test Structure,
Earthquake Effects on Reinforced Concrete Structures U.S.- Japan Research,, Publication SP-84,
ACI, pp. 203-239, 1985.
5. Takeda, T., Mete A. Sozen, N. Norby Nielsen (1970), Reinforced Concrete Response to Simulated
Earthquake, Journal of Structural Division, Vol. 96, ST 12.

You might also like