Você está na página 1de 22

CAPACITY DESIGN, A CONCEPT TO ENSURE

SEISMIC RESISTANCE OF BUILDING STRUCTURES∗ )

By :

Prof.Dr.Ir. Wiratman Wangsadinata

Professor of Structural Engineering


Department of Civil Engineering, ITB

President Director PT. Wiratman & Associates

ABSTRACT

This paper describes the capacity design principles in seismic resistant design of structures along with the
use of the Load Resistance Factor Design (LRFD) method. As one form of performance based seismic
design, the value of the ductility factor of the structure may be selected according to the need, ranging from
the value for the full elastic up to that for the full ductile structure. The latest version of the seismic zoning
map of Indonesia and the elastic as well as reduced response spectra for each zone are given. In the
implementation of the strong column weak beam concept it is shown, that in arriving at the nominal forces
originating from the actual existing flexural capacities, the equilibrium of forces may be considered either at
the state of near collapse of the structure or at the onset of first yielding. Finally, the treatment of the 3D
building structure as an assemblage of various subsystems with different ductility levels is given.

1. INTRODUCTION

Capacity Design is a concept or a method of designing flexural capacities of critical


member sections of a building structure based on a hypothetical behaviour of the structure
in responding to seismic actions. This hypothetical behaviour is reflected by the
assumptions that the seismic action is of a static equivalent nature increasing gradually
until the structure reaches its state of near collapse and that plastic hinging occurs
simultaneously at predetermined locations to form a collapse mechanism simulating
ductile behaviour.
The actual behaviour of a building structure during a strong earthquake is far from
that described above, with seismic actions having a vibratory character and plastic hinging
occurring rather randomly. However, by applying the Capacity Design concept in the
design of the flexural members of the structure, it is believed that the structure will possess
adequate seismic resistance, as has been proven in many strong earthquakes in the past.
Therefore, many seismic design codes around the world, including the draft of the new
Indonesian Seismic Design Code SNI-1726, adopt the Capacity Design concept as a
normative requirement.
A feature in the Capacity Design concept is the ductility level of the structure,
expressed by the displacement ductility factor or briefly ductility factor. This is the ratio
of the lateral displacement of the structure due to the Design Earthquake at near collapse
and that at the point of first yielding. The value of this ductility factor may be selected by

∗)
Presented to the First National Conference on Earthquake Engineering, Institut Teknologi Bandung,
Nov. 4-5, 1999.

1
the Owner or the Designer of the building according to the need, ranging from the value
for the full elastic up to that for the full ductile structure. This can be considered as one
form of performance based seismic design, where the targeted performance of the
structure is quantified through the selected ductility factor.

2. DESIGN EARTHQUAKE

For normal buildings the Design Earthquake is an earthquake causing its structure to
reach a state of near collapse having a peak ground acceleration with a 10% probability of
exceedence in 50 years. This 50 years time span is the assumed life time of normal
buildings. With this criteria, based on probability theorem the Design Earthquake has a
mean return period of about 500 years.
For buildings of post-earthquake importance (for example hospitals), the probability
of exceedence of the earthquake causing collapse in 50 years should be lowered (say to
5%), implying a mean return period of the earthquake longer than 500 years. For
monumental buildings (for example monuments and very tall buildings), which must be
preserved or must remain usable for a very long time, apart from the fact that they must
have a life time longer than 50 years (say 100 years), the probability of exceedence of the
earthquake causing collapse should also be lowered, implying a mean return period of the
earthquake of much longer than 500 years. For both cases it is convenient to consider the
same Design Earthquake as for normal buildings, but its action is multiplied by an
importance factor I larger than 1 (I > 1).
The fact that certain buildings have life times shorter than 50 years (for example
simple dwelling houses) should certainly be accounted for, even if the probability of
exceedence of the earthquake causing collapse during their shorter life time is taken the
same as for normal buildings (10%). For this case the same Design Earthquake as for
normal buildings could be used, but its action is multiplied by an importance factor I less
than 1 (I < 1).

3. THE STANDARD 2D BUILDING STRUCTURE

3.1. DUCTILITY PARAMETERS

The standard 2D building structure is a hypothetical one, which is modelled as a


structure that if the action of the Design Earthquake on it reaches a certain level, it is
capable of simultaneous first yielding (plastic hinging) at all beam ends and all column
bases, producing a perfect collapse mechanism as shown in Figure 1(a). It is said that the
structure meets the so called strong column weak beam concept. Before this level of action
is reached, the structure behaves perfectly elastic. Although the structure has reached a
collapse mechanism, it can still absorb some additional seismic load due to the strain
hardening effects of the yielding steel in the rotating plastic hinges, increasing the yield
moment capacities while lateral displacement progresses. This structural performance can
best be demonstrated by inspecting its load-deflection diagram (V-δ diagram) as shown in
Figure 1(b) and 1(c), where the load is represented by the base shear V and the deflection
by that of the top of the structure δ.

2
Consider first the case of the structure, where the yield moment capacities at all
beam ends and all column bases have been given a very high value, so that imposed by the
gradual effect of the Design Earthquake, no plastic hinges are developed in the structure,
causing it to behave fully elastic. The load-deflection diagram is a straight line, where the
maximum elastic load absorbed by the structure before collapse Ve is associated with a
maximum deflection δ m as shown in Figure 1(b) and 1(c).
δ
Fn

Fk

F2 zk
F1
V

(a)
V V
Ve Ve

½V e

V cap,i V cap,i f oiV yi


foi Vyi
V y,i V y,i
V cap foV y V cap f oV y
Vy Vy

0 δ y δ yi δm δ 0 δy δ yi ½δ m δm δ
(b) (c)
Figure 1. Collapse mechanism and load-deflection diagrams of the standard 2D building
structure.

Now the yield moment capacities at all beam ends and all column bases are reduced
to a minimum value, changing the structure into a full ductile one. It means that it can
produce the maximum possible post-elastic deflection without collapse. The load-
deflection diagram is the same straight line up to the load level Vy , at which simultaneous
first yielding occurs at all beam ends and all column bases, initiating the formation of its
collapse mechanism. Vy is referred to as the first yield load, which is associated with the
first yield deflection δ y as shown in Figure 1(b) and 1(c). Applying the constant
displacement rule for the full elastic and the full ductile structure, the following
relationship can be expressed :

3
δ m Ve
µm = = ………. (1)
δy Vy

where µm is referred to as the displacement ductility factor or briefly the ductility factor of
the full ductile structure, which is the maximum that the standard 2D building structure
can achieve.
As the structure in its collapse mechanism under seismic action continues to deflect
laterally, accompanied by continued rotation of all plastic hinges, strain hardening effects
of the yielding steel in the plastic hinges cause the yield moment capacity of every plastic
hinge to increase gradually from its nominal value My at the point of first yielding up to its
maximum value before collapse Mcap. This phenomena in turn causes the total seismic
load absorbed by the structure to increase gradually from its nominal value Vy at the point
of first yielding up to its maximum value before collapse Vcap. In this case the maximum
yield moment capacity Mcap of a plastic hinge and the maximum total seismic load Vcap
absorbed by the structure before collapse can be expressed as follows :

Mcap = fo My ………. (2)

Vcap = fo Vy ………. (3)

where fo is the strain hardening overstrength factor (see Figure 1(b) and 1(c)). The post-
elastic load-deflection diagram may be idealized in several ways. In Figure 1(b) it is
idealized as a sloping line, while in Figure 1(c) as a horizontal line of length ½ δ m (the
yield plateau) followed by a sloping line. It will be shown that it doesn’t matter too much
which idealization is used, as it will give minor difference in results.
As a general case a structure will now be considered, where yield moment capacities
at all beam ends and all column bases have a general value higher than those of a full
ductile structure. Such structure is referred to in this paper as a partially ductile one,
having a ductility factor µi, first yield load Vyi, first yield deflection δ yi, strain hardening
overstrength factor foi, maximum yield moment capacity of a plastic hinge Mcap,i and
maximum total seismic load Vcap,i (see Figure 1(b) and 1(c)), respectively according to the
following expressions :

δm V
µi = = e ………. (4)
δ yi Vyi

Mcap,i = foi Myi ………. (5)

Vcap,i = foi Vyi ………. (6)

Regarding the value of foi, this is a function of µi and depends on the value of µm and fo
and also the idealization of the post-elastic V-δ diagram. Using the idealization as a
sloping line with equal slope as shown in Figure 1(b), it can be shown that the following
relationship applies :

(µi − 1) (f o − 1)
f oi = 1 + ………. (7)
µ m −1

4
and using the idealization as a horizontal line followed by a sloping line with equal slope
as shown in Figure 1(c), it can be shown that the relationship is as follows :

- for 1 < µi < 2 :


foi = 1 ………. (8)

- for 2 < µi < µm


f oi = 1 +
(µi − 2 ) ( fo − 1)
………. (9)
µm − 2

According to the draft of the new Indonesian Seismic Design Code SNI-1726, for a
full ductile structure µm = 5,3 and fo = 1,25. Substituting these values for µm and fo in
eq.(7) and (9) the strain hardening overstrength factor of a partially ductile structure for an
idealized post-elastic V-δ diagram as a sloping line with equal slope becomes :

foi = 0,94 + 0,06 µi ………. (10)

and for an idealized post-elastic V-δ diagram as a horizontal line followed by a sloping
line with equal slope becomes

- for 1 < µi < 2 :


foi = 1 ………. (11)

- for 2 < µi < 5,3


foi = 0,85 + 0,08 µi ………. (12)

In Table 1 values of foi for variable µi, for the whole range of structures from the full
elastic up to the full ductile ones according to eq.(10), (11) and (12) are listed.

Table 1. Ductility and Design Parameters for the standard 2D Building Structures

foi Ri fi ’ Ki Ω Si Ωi
Level of
µi Eq.(10) Eq.(11) Eq.(16) Eq.(19) Eq.(20) Eq.(36) Eq.(34)
Performance
& (12)

Full elastic 1.0 1.00 1.00 1.6 1.6 5.3 1.00 1.6

1.5 1.03 1.00 2.4 1.6 3.5 1.09 1.7


2.0 1.06 1.00 3.2 1.7 2.7 1.17 1.9
2.5 1.09 1.05 4.0 1.7 2.1 1.26 2.0
Partially 3.0 1.12 1.09 4.8 1.8 1.8 1.35 2.1
ductile 3.5 1.15 1.13 5.6 1.8 1.5 1.44 2.3
4.0 1.18 1.17 6.4 1.9 1.3 1.52 2.4
4.5 1.21 1.21 7.2 1.9 1.2 1.61 2.6
5.0 1.24 1.25 8.0 2.0 1.0 1.70 2.7

Full ductile 5.3 1.25 1.25 8.5 2.0 1.0 1.75 2.8

5
From Table 1 it can be seen, that from the values of foi shown indeed it doesn’t matter too
much how the post-elastic V-δ diagram is idealized, since both idealizations give
practically the same values. In this case, the draft of the new Indonesian Seismic Design
Code SNI-1726 uses the sloping line as the idealized post-elastic V-δ diagram as shown in
Figure 1(b) and expressed by eq.(10).

3.2. DESIGN PARAMETERS

For the strength design purposes of the members of the standard 2D building
structure using the Load and Resistance Factor Design (LRFD) method adopted in the
draft of the new Indonesian Seismic Design Code SNI-1726, the total design nominal
seismic load Vni for a partially ductile structure with a certain ductility factor µi must be
defined, in such a way lower than the first yield load Vyi so that the total design
overstrengths associated with excessive material strength and excessive loading and load
combinations encountered in practice are incorporated. Introducing a total design
overstrength factor f1 as a multiplication of the material overstrength factor ΩM and the
load overstrength factor Ω D (see Figure 2), i.e. :

f1 = Ω M ΩD ………. (13)

the total design nominal seismic load Vni can be expressed as

Vyi
Vni = ………. (14)
f1

If there are no overstrengths at all, theoretically Ω M = γm = 1.15 (= 1/φ) and Ω D = γs =


1.05 in accordance with the LRFD principles defined in the draft of the new Indonesian
Seismic Design Code SNI-1726, where γm is the material factor (φ = capacity reduction
factor) and γs is the overall load factor. Therefore, theoretically f1 = 1.15 x 1.05 = 1.2. In
practice there are always deviations in member section properties due to excess in the
number of rebars or oversize in steel section or dimensions, causing increase in the value
of the material overstrength factor. There are also always deviations in the loading due to
excess in the loading itself or difference in load combinations, causing increase in the
value of the load overstrength factor. For the Indonesian condition it is considered
appropriate and realistic to adopt an overall value for the total design overstrength factor
of

f1 = 1.6 ………. (15)

as defined in the draft of the new Indonesian Seismic Design Code SNI-1726.
The total design nominal seismic load Vni for a partially ductile structures with a
certain ductility factor µi can now be expressed in the maximum elastic load Ve by
introducing a total reduction factor Ri as follows :

Ri = µi f1 ………. (16)
so that

6
Ve
Vni = ………. (17)
Ri

In Table 1 values of Ri for variable µi for the whole range of structures from the full
elastic up to the full ductile ones according to eq.(16) are listed.
If the maximum possible first yield load V yi and the corresponding maximum
possible total seismic load Vcap,i for a partially ductile structure with a certain ductility
factor µi is given, computed based on the actual dimensions and existing number of rebars
or actual cross-sectional properties of the structural members, the maximum possible total
design nominal seismic load V ni is to be computed from the following expression :

Vcap,i f oi V yi V yi
V ni = = = ………. (18)
f i' f i' f1

where (see Figure 2) fi ’ is the total capacity overstrength factor defined as follows :

f i' = f oi x f1 = 1,6 f oi ………. (19)

V
Ve

µi
f oiVyi
Ri Vcap,i
Vyi f oi
ΩM f’
f1
ΩD Vni

0 δyi δm δ

Figure 2. Load-deflection diagram of the standard 2D building structure


and its design parameters.

In Table 1 values of fi ’ for variable µi for the whole range of structures from the full elastic
up to the full ductile ones according to eq.(19) and eq.(10) are listed. In eq.(18) V ni , Vcap,i
and V yi must be seen as general forces, so that they may represent moments, shears or
axial forces. From eq.(18) it is apparent, that to determine V ni there is an option to avoid

7
the use of the strain hardening overstrength factor foi. Furthermore, in capacity design
consideration, it is obvious that V ni computed according to eq.(18) need not be taken
more than that produced by the Design Earthquake under elastic condition of the structure.
In practice it is customary and in fact defined in the draft of the new Indonesian
Seismic Design Code SNI-1726 that the seismic action upon a partially ductile structure
with a certain ductility factor µi is obtained from the action attained by a full ductile
structure, by multiplying it by a structural type factor Ki. It is obvious that Ki is defined by
the following expression :

µm 5,3
Ki = = ………. (20)
µi µi

With µm = 5.3, in Table 1 values of Ki for variable µi for the whole range of structures
from the full elastic up to the full ductile ones according to eq.(20) are listed.
If the total reduction factor for a full ductile structure is denoted by Rm, which from
Table 1 has a value of 8.5, the structural type factor Ki may also be expressed as

R m 8.5
Ki = = ………. (21)
Ri Ri

3.3. SEISMIC ZONING AND RESPONSE SPECTRA

The latest version of the seismic zoning map of Indonesia (adopted by the working
group on seismic zoning on October 1, 1999) is shown in the attachment. The
differentiation of the zones is based on the Peak Base Acceleration (PBA) of the Design
Earthquake with a 10% probability of exceedence in 50 years, thus with a mean return
period of about 500 years. As can be seen, Indonesia is divided into 6 Seismic Zones, each
zone with an average Peak Base Acceleration as shown in Table 2.

Table 2. Peak Base Acceleration and Peak Ground Acceleration


for the seismic zones of Indonesia

Peak Base Peak Ground Acceleration (‘g’)


Zone Accel.
(‘g’) Hard Soil Medium Soil Soft Soil Special Soils

1 0.025 0.030 0.045 0.075 Site specific


evaluation
2 0.100 0.120 0.180 0.250 required

3 0.175 0.220 0.270 0.367

4 0.225 0.280 0.350 0.420

5 0.275 0.340 0.391 0.442

6 0.325 0.400 0.440 0.480

8
In Table 2 the respective Peak Ground Acceleration (PGA) for each seismic zone is
indicated for 3 soil categories, namely Hard Soil, Medium Soil and Soft Soil, respectively
equivalent to S2 , S3 and S4 soil categories according to UBC 1997. The PGA value for
Hard Soil in Table 2 is approximately 1.25 times the respective PBA value in each zone.
Special soils in Table 2 refer to very soft soils, soils with high liquefaction potentials, soils
with high organic content, etc., which require site specific evaluation. To identify the soil
category, Table 3 gives the definition, where the soil category is defined depending on the
weighted average N-SPT value of the soil and the depth of the hard layer. In this relation,
the hard layer is defined as the layer below the surface, which has an N-SPT value of at
least 50, provided that no other deeper layers are present with lower N-SPT values. Here
the N-SPT value is used as the soil parameter, because it can be easily found in any
standard geotechnical investigation report, thus facilitating designers.

Table 3. Definition of soil category

Soil category
Depth of hard layer
(m)
Hard Soil Medium Soil Soft Soil

5 N > 20 10 < N < 20 N < 10

10 N > 30 20 < N < 30 N < 20

15 N > 40 30 < N < 40 N < 30

> 20 N > 50 40 < N < 50 N < 40

Instead of N-SPT values, various codes specify shear wave velocities as the criteria
for defining the soil category. In this respect, Figure 3 gives the empirical correlation
between N-SPT value and shear wave velocity Vs, useful for comparison purposes.
The correlation functions shown are as follows :
100
80
60
40

Ohsaki, Iwasaki
20
Vs (m/det)

10
8 Author Okamoto
6
4

1
1 2 4 6 8 10 20 40 60 80100
N
Figure 3. Empirical correlation between N-SPT value and shear wave
velocity Vs (dots are after Okamoto).

9
- after Okamoto :
N = 0.00035 Vs2 ………. (22)
- after Ohsaki & Iwasaki :
N = 0.000015 Vs2.5 ………. (23)
- proposed by the author as an average function :
N = 0.000075 Vs2.25 ………. (24)

The Peak Base Acceleration of the Design Earthquake multiplied by the considered
Importance Factor I is generally used as the input maximum base rock acceleration in site
specific response studies or in soil-structure interaction problems, where seismic waves
travelling from the base rock to the ground surface undergo amplifications depending on
the soil properties. The Peak Ground Acceleration of the Design Earthquake multiplied by
the considered Importance Factor I is generally used as the input maximum ground surface
acceleration in time history non-linear dynamic response analyses of the building structure
to verify its seismic resistance against the Design Earthquake at near collapse. If time
history dynamic response analysis is used for design purposes, the input maximum ground
acceleration due to the Design Earthquake must be reduced by the Reduction Factor Ri for
the respective level of ductility in line with eq.(17), taking into account also the considered
Importance Factor I. The analysis remains therefore in the elastic range.
For design purposes response spectrum modal analysis is more frequently used,
applying the standard response spectra of the Design Earthquake. Attached are 2 sets of
standard response spectra for each Seismic Zone for the 3 different soil categories. The
first set is the standard elastic response spectra Sa-T for a full elastic structure (Ri = 1, µi =
1, Ki = 5,3), the second one is the standard reduced response spectra C-T for a full ductile
structure (Ri = 8.5, µi = 5.3, Ki = 1), called the nominal response spectra. The reduction of
the elastic response spectra by the total reduction factor for a full ductile structure Rm =
8.5 to become the nominal response spectra is in line with eq.(17). It should be noted, that
for zero natural period (T = 0) for building structures of normal importance (I = 1) the
response acceleration of the elastic structure becomes identical with the PGA of the
Design Earthquake, while the response acceleration of the full ductile structure becomes
identical with the reduced PGA (reduced by Rm = 8.5) of the Design Earthquake.
Knowing the fundamental natural period of vibration of the standard 2D building
structure T1 , the base shear coefficient Ci for T = T1 is obtained from the nominal response
spectra. The nominal seismic base shear for the building with an Importance Factor I and
a structural type factor Ki may be expressed as follows :

Vi = Ci I Ki Wt ………. (25)

where Wt is the total weight of the building, including an appropriate portion of the live
load. Eq.(25) is based on the theory of modal analysis, which indicates that in general the
response of a building structure is dominantly governed by the response of its first mode.
The nominal seismic base shear Vi is further distributed into nominal earthquake
loads, each acting at the center of mass of each floor according to the following expression

10
Wk z k
Fki = m
Vi ………. (26)
∑ Wk z k
k =1

where k is the floor number, m the top floor number, zk the height of floor k from the base
and Wk the weight of floor k, including an appropriate portion of the live load (see Figure
1(a)). Eq.(26) is derived from the theory of modal analysis in which the first mode shape is
simplified into a straight line and the response of the structure represented by the response
of its first mode only. The nominal earthquake loads according to eq.(26) may then be
considered as equivalent static seismic loads, against which the structure may be analysed
for its internal forces. In the LRFD method these nominal seismic forces are then
combined with the other nominal forces, i.e. due to dead load and live load, after each has
been multiplied by its respective load factor, resulting in the ultimate load acting on the
structural member.

3.4. STRONG COLUMN WEAK BEAM CONCEPT

To show how the strong column weak beam concept is implemented through the
LRFD method, consider the standard 2D building structure shown in Figure 1(a) at its
state of near collapse. An enlarged beam to column connection of the structure is shown in
Figure 4(a), while in Figure 4(b) and 4(c) the load and internal force configuration are
shown for a beam and a joint respectively. The structure is supposed to be partially ductile
with a structural type factor Ki.
In the first place, the beam must not fail in shear at the state of near collapse of the
structure. Denoting the nominal value of the yield moments at both beam ends by Myi and
M 'yi and the respective maximum values at near collapse by Mcap,i and M 'cap,i , all of which
computed based on the actual section properties, following eq.(18) at the nominal level
those moments produce nominal shears in the beam Q E,n as follows (see Figure 4(b)) :

Q E,n =
M cap,i + M 'cap,i
=
(
f oi M yi + M 'yi ) = M yi + M 'yi ………. (27)
f i' l ' f i' l ' f1 l '

where foi and f i' are the strain hardening and the total capacity overstrength factores, both
for the existing level of ductility, f1 = 1.6 is the total design overstrength factor and l’ is
the clear span of the beam.
It is obvious that the nominal shears in the beam according to eq.(27) need not be
taken more than those produced by the Design Earthquake under elastic condition of the
structure. If in this relation Q E,n are the nominal shears in the beam produced by the
action of the nominal earthquake loads under partially ductile condition of the structure
expressed by its Ki factor, the said limitation can be expressed as follows :

5.3
Q E,n ≤ Q E,n ………. (28)
Ki

11
plastic
plastic hinge
hinge

plastic
l’ hinge

l
(a)
M’ E
M’ cap,i

M cap,i M cap,i

M’cap,i
QE,n QE,n
QD,n; QL,n QD,n; QL,n
l’ ME
(b) (c)

Figure 4. Strong column weak beam concept implemented


through the LRFD method.

In the LRFD method the combined factored nominal forces are then considered to
produce the ultimate condition :

Q u = γ D Q D,n + γ L Q L, n + γ E Q E, n ………. (29)

where QD,n and QL,n are the shears due to the nominal dead load and nominal live load
respectively, while γD and γL are the corresponding load factors and γE is the load factor
for seismic load. According to the draft of the new Indonesian Seismic Code SNI-1726, γD
= γL = γE = 1.05.
Now the beam plastic hinging will be considered. Since plastic hinges must not
developed in the column (strong column weak beam), the sum of the yield moment
capacities of the columns must at least be equal to the sum of the yield moment capacities
of the beams meeting at the joint. Denoting the yield moment capacity of a column (say
the lower column) at the nominal level M E,n , applying moment distribution principles
and following eq.(18), the following expression for M E,n can be written (see Figure 4(c)):

12
h 'c M cap,i + M 'cap,i
M E,n = α c ωd
hc f i'

= αc
h 'c
ωd
(
f oi M yi + M 'yi ) ………. (30)
hc f i'
h 'c M yi + M 'yi
= αc ωd
hc f1

where α c is the moment distribution factor for the considered (lower) column, hc ’/hc is the
ratio of the clear height to the total (theoretical) height of the column, ωd is a dynamic
amplification factor to account for the increased response of the column due to beam
plastification and f1 = 1.6 is again the total design overstrength factor. This column
moment is of course associated with a corresponding axial force, which is not shown here.
It is obvious that the nominal yield moments of the column according to eq.(30)
need not be taken more than those produced by the Design Earthquake under elastic
condition of the structure. Denoting ME,n as the nominal yield moment in the column
produced by the action of the nominal earthquake loads under partially ductile condition
of the structure expressed by its Ki factor, the said limitation can be expressed as follows :

5 .3
M E,n ≤ M E,n ………. (31)
Ki

In the LRFD method the combined factored nominal moments are then considered
to produce the ultimate condition :

M u = γ D M D, n + γ L M L , n + γ E M E , n ………. (32)

where MD,n and ML,n are the moments due to the nominal dead load and nominal live load
respectively, while γD and γL are the corresponding load factors and γE is the load factor
for seismic load. As mentioned before according to the draft of the new Indonesian
Seismic Design Code SNI-1726, γD = γL = γE = 1.05. The corresponding ultimate axial
force is not shown here.
Eqs.(27) and (30) (also eq.(18)) clearly show that in arriving at the nominal forces
originating from the actual flexural capacities based on existing cross sectional properties,
the equilibrium of forces may be considered either at the state of near collapse of the
structure (as indicated in the present Indonesian Concrete Code SNI-03-2847-1992) or at
the onset of first yielding in the structure.

3.5. STRUCTURAL OVERSTRENGTH AND LOADS ON THE FOUNDATION

In actual condition the standard 2D building structure designed for its flexural
strength based on the Capacity Design concept, will not behave as described previously.
Plastic hinge formation will not occur simultaneously at all beam ends and all column
bases, making the structure stronger than assumed in design. These are due to various
factors, such as redistribution of internal member forces in the highly redundant structure,

13
amount of steel rebars exceeds that required by the seismic force, member oversize,
minimum requirements on proportioning and detailing, effect of on-structural elements,
etc. The V-δ diagram for a partially ductile structure looks like shown in Figure 5. After
reaching the point of first yielding at the load Vyi, the curve follows a steep ascending
curve until the maximum deflection δ m is reached at the point of near collapse with the
actual maximum total seismic load absorbed by the structure :

Vai = Ω Si Vyi ………. (33)

where Ω Si is referred to as the structural overstrength factor for the partially ductile
structure.

V
Ve

ΩSiVyi
Vai
ΩSi
Ωi
Vyi
f1
Vni

0 δyi δm δ

Figure 5. Actual load-deflection diagram for a partially ductile structure.

Introducing Ω i as the total structural overstrength factor according to the following


expression :

Ω i = Ω Si . f1 = 1.6 Ω Si ………. (34)

the actual maximum total seismic load can be expressed in the total design nominal
seismic load Vni as follows :

Vai = Ω i Vni ………. (35)

If a full ductile structure with a ductility factor µm is known to possess a structural


overstrength factor Ω S, then applying the equal slope principle, the structural overstrength
factor for a partially ductile structure with a ductility factor µi may be expressed as follows

14
(µi − 1) (ΩS − 1)
Ω Si = 1 + ………. (36)
µm −1

The draft of the new Indonesian Seismic Design Code SNI-1726 defines Ω S = 1.75 and µm
= 5.3, so that for the whole range of structures from the full elastic up to the full ductile
ones, the structural overstrength factor Ω Si and the total structural overstength factor Ω i
computed according to eqs.(36) and (34) are as listed in Table 1.
The existence of the structural overstrength in building structures is an important
characteristic, recognized as a major factor in the survival of buildings during severe
earthquakes in the past. However, such survival will not happen if the foundation or
substructure has failed earlier. It is therefore important to access the structural
overstrength of a building structure thoroughly, so that its foundation or substructure can
be designed accordingly. This is part of the capacity design principle.
Representing the seismic loads on the foundation or substructure exerted by the
superstructure by the base shear, there are 3 loading conditions that have to be considered :

- the actual loads under actual structural overstrength :

Vai = Ω i Vni ……….. (I)

- the maximum loads generated at the point of near collapse under the action of the
maximum yield moment capacities of all plastic hinges at the base :

V cap,i = f oi V yi ………. (II)

- the actual loads produced by the Design Earthquake under elastic condition of the
superstructure :

5.3
Ve = f1 Vni ………. (III)
Ki

It should be noted in this respect, that eqs.(I), (II) and (III) must be seen as general forces,
so that they are representing moments, shears and axial forces. The governing loads on
the foundation or substructure are the highest between the ones expressed by eqs.(I) and
eq.(II), but in any case they need not be taken more than that expressed by eq.(III).
According to the draft of the new Indonesian Seismic Design Code SNI-1726, under
the action of seismic loads the foundation or substructure must remain elastic. Therefore,
for checking purposes of the adequacy of the foundation or substructure the loads acting
on it are those expressed directly by eqs.(I), (II) or (III), but for design purposes based on
LRFD principles they have to be reduced by the total reduction factor for the elastic
structure of the foundation or substructure R = f1 = 1.6. Finally, the complete loading on
the foundation or substructure of course include the loads from the superstructure due to
dead load and live loads, which are not shown here.

15
4. THE GENERAL 3D STRUCTURE

4.1. ASEMBLAGE OF SUBSYSTEMS

In actual condition a building structure has always a 3D configuration. Especially to


resist seismic action, the 3D character of the structure is most essential to cope with the
fact, that the direction of the earthquake action is arbitrary. Since for several directions the
3D system generally does not consist of only one sort of subsystem, but may consist of
several subsystems with several levels of ductility, it is therefore important to know what
the overall ductility level is of the 3D structure as a whole.
The different ductility levels of 2D subsystems can best be categorized by
comparing their ductility and design parameters with those of the full ductile standard 2D
building structure (open frame) discussed in the previous chapter, which has a ductility
factor of µm = 5.3 and corresponding total reduction factor of Rm = 8.5. Denoting the
maximum achievable ductility factor of a subsystem by µms and the corresponding total
reduction factor by Rms (= 1.6 µms), the subsystem may be considered to have a structural
type factor given by :

5.3 8.5
Ks = = ………. (37)
µ ms R ms

For some 2D structural subsystems their µms, Rms and corresponding Ks values are listed
in Table 4. Assuming that those subsystems fit well into the spectrum of standard 2D
building structures shown in Table 1, knowing the value of their ductility factor µms, their
corresponding overstrength factors fos, fs ’, Ω Ss and Ω s may be computed from the
respective equations, the result of which listed in Table 4. This table can be extended as
necessary to cater for other types of structural subsystems when their µms or Rms value
becomes available. According to the need, for a certain subsystem a lower value of its
ductility factor than that shown in Table 4 may be used, even until as low as that for a full
elastic structure (µ s = 1). In that case Table 1 may be used to find its corresponding
ductility and design parameters.

Table 4. Ductility and Design Parameters for some 2D structural subsystems.

Type of 2D Subsystem µms Rms Ks fos f’s Ω Ss Ωs


Eq.(10) Eq.(19) Eq.(36) Eq.(34)

Chimneys 1.8 2.8 3.0 1.05 1.7 1.14 1.8

Diagonally braced frames 2.1 3.4 2.5 1.07 1.7 1.19 1.9

Prestressed concrete 3.8 6.1 1.4 1.17 1.9 1.49 2.4


Frames

Cantilever shear walls 4.4 7.0 1.2 1.20 1.9 1.59 2.5

Open frames, coupled 5.3 8.5 1.0 1.25 2.0 1.75 2.8
shear walls

16
Knowing the value of th e Ks-factor for each subsystem, the problem is how to
determine the representative value of the K-factor of an assemblage of several subsystems.
Provided that the difference between the Ks-factor of the various subsystems is not more
than 0.5, a method of determining this overall K-factor is to calculate it as the weighted
average, using the base shear resisted by each subsystem as the weighing factor. As an
example, a building structure with an orthogonal axis system will be considered. The
weighted average of the structural type factor for seismic action in the x-direction is

Σ K xs Vxs
Kx = ………. (38)
Σ Vxs

and for seismic action in the y-direction is

Σ K ys Vys
Ky = ………. (39)
Σ Vys

so that the overall structural type factor becomes

K x Vx + K y Vy
K = ………. (40)
Vx + Vy

where Vxs and Vys are obtained from a 3D analysis of the structure subjected to earthquake
action in x-direction and y-direction respectively.

4.2. EQUIVALENT STATIC AND DYNAMIC ANALYSIS

Equivalent static analysis using equivalent static seismic loads as defined from
eq.(25) and (26) may be applied to regular building structures. Structures rectangular in
plan are generally regular. No vibration analysis is required for such a structure, but if it is
carried out its first and second natural modes of vibration will be dominant in translation
in the two perpendicular directions of its geometrical axes. Therefore, the structure
behaves almost like a 2D structure in the direction of each of its geometric axes, which
naturally is selected as the direction of the seismic action. For the first seismic direction,
the base shear may be determined from eq.(25) for the first natural period, while for the
second seismic direction perpendicular to the first, the base shear may be determined from
eq.(25) for the second natural period. Since no vibration analysis is required, the natural
period for each of the seismic directions is computed using the well known Rayleigh’s
formula on the basis of the static deflection of the structure. As there will be always
eccentricities between the center of mass and the center of rotation on each floor, 3D static
analyses should preferably be carried out.
For irregular building structures 3D dynamic response analysis must be applied.
Structures with non-symmetrical shape in plan are generally irregular. It is required by the
draft of the new Indonesian Seismic Design Code SNI-1726 that any building structure
must have at least its first (fundamental) natural mode of vibration dominant in translation.
The second and higher modes of vibration may already be dominant in torsion. For the
direction of this first mode translation, a reference base shear V can be calculated from

17
eq.(25). The draft code then requires, that any base shear of the irregular structure for any
other considered seismic direction must not be taken less than 0.8 V.
Non-linear static as well as dynamic analyses are generally not required for design
of normal building structures. But for structures containing subsystems of unknown levels
of ductility, static push-over or non-linear time history dynamic response analyses must be
performed, to verify the correctness of the assumed ductility level, thus to ensure adequate
seismic resistance of the structure against the Design Earthquake at near collapse.

4.3. CAPACITY DESIGN

In a 3D building structure comprising of several subsystems of different levels of


ductility, each individual subsystem behaves similarly like a 2D structure. Therefore, the
principles of Capacity Design described previously for the standard 2D building structure
may be applied similarly to each of those individual subsystems.
Also in a 3D building structure must its foundation or substructure as a whole
remain intact and elastic during the action of the Design Earthquake. Therefore, eqs.(I),
(II) and (III) must be considered for every direction of the seismic action considered. In
this connection it should be noted, that the governing direction of seismic action for the
superstructure may not necessarily be the same as that for the foundation or substructure.
The most critical direction of seismic action on the foundation or substructure as a whole
must be investigated separately.

5. CONCLUSION

From the previous discussion, the following can be concluded :

1. The Capacity Design is a concept of designing flexural capacities of structural


member sections, based on a hypothetical behaviour of the structure, however it
ensures the structure from failure during strong earthquakes.

2. Since in the Capacity Design the level of ductility of the structure may be selected
according to the need, this can be considered as one form of performance based
seismic design, where the targeted performance is quantified through the selected
ductility factor.

3. In the implementation of the strong column weak beam concept, the equilibrium of
forces may be considered either at the state of near collapse of the structure or at the
onset of first yielding.

4. As part of the Capacity Design principle, foundation or substructures must not fail
earlier than the superstructure, so that it is important to access the actual loading
from the superstructure, taking into account all possible overstrengths present in the
superstructure.

18
REFERENCES

1. Dep. of Public Works (1999) : “Indonesian Seismic Code for Buildings” (in
Indonesian), SNI-1726, 2nd draft revision, 1999.

2. Dep. of Public Works (1992) : “Indonesian Concrete Code for Buildings” (in
Indonesian), SNI-03-2847-1992, Yayasan Badan Penerbit PU, 1989.

3. Firmansyah, J. (1999) : “Development of Seismic Hazard Map for Indonesia”,


Seminar HAKI 1999, Industri Konstruksi Indonesia Menghadapi Era Millennium
Ketiga, Jakarta, 8 Sept., 1999.

4. ISO (1999) : “Bases for Design of Structures, Seismic Actions on Structures”,


ISO/TC 98/WG-1 N50, ISO/CD 3010 (Rev.), October 1, 1999.

5. NEHRP (1997) : “NEHRP (National Earthquake Hazards Reduction Program)


Recommended Provisions for Seismic Regulations for New Buildings and Other
Structures”, Part 1, Part 2, Building Seismic Safety Council, Federal Emergency
Management Agency, Washington D.C., 1997.

6. Wangsadinata, W. (1996) : “Some Fundamental Criteria of Earthquake Loading on


Building Structures”, International Workshop on Regional Harmonization of
Loading Specifications, Melbourne, Australia, Nov. 20 – 22, 1996.

7. Wangsadinata, W. (1997) : “Some Fundamental Parameters in Design Codes for


Building Structures”, International Workshop on Harmonization in Performance
Based Building Structural Design in Countries surrounding Pacific Ocean, Tsukuba,
Japan, Dec. 1 – 3, 1997.

8. Wangsadinata, W. (1998) : “Fundamental Parameters for the Seismic Design of


Building Structures in Indonesia”, Journal Teknik Sipil, Fakultas Teknik, Untar,
No.1, Tahun ke-IV, Maret 1998.

9. Wangsadinata, W. (1998) : “The Draft of the New Indonesian Seismic Code”,


International Seminar and Workshops on International Standards on Loadings and
Structural Design, Singapore, July 6 – 8, 1998.

10. Wangsadinata, W.; Surahman, A.; Sidjabat, H.R. (1999) : “Report on the Revision of
the Indonesian Seismic Code”, APEC Workshop & Seminar on the International
Framework for Structural Design, Shenzhen, China, March 22 – 25, 1999.

11. Wangsadinata, W,; Mangkusubroto, S.; Surahman, A.; Djajaputra, A.; Kertapati, E.;
Toha, F.X.; Jitno, H.; Sengara, I.W.; Firmansyah, J.; Irsyam, M.; Puspito, N.;
Rudianto, S.; Sutjipto, S.; Najoan, T. (1999) : “Development of Seismic Hazard Map
for Indonesia”, Interim Result, APEC Workshop & Seminar on the International
Framework for Structural Design, Shenzhen, China, March 22 – 25, 1999.

19
20
SPEKTRUM RESPONS ELASTIK Sa (R=1, µ=1, K=5.3)

Sa Wilayah 1 Sa Wilayah 2
0.625 0.500
Sa = (lunak)
T
0.270
0.500 Sa = (sedang)
T
0.450
0.120
Sa = (keras)
T
0.150
Sa = (lunak)
T 0.300
0.270
0.068 0.250
Sa = (sedang) 0.250
T
0.188 0.167
0.030 0.180
0.150 Sa = (keras)
T 0.135
0.120 0.120
0.113 0.090
0.075 0.068 0.075 0.060
0.050 0.040
0.045 0.030 0.034
0.015 0.023
0.030 0.010
0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0
T T

Sa Wilayah 3 Sa Wilayah 4
1.050

0.840
Sa = (lunak)
0.875 0.840 T
0.734
Sa = (lunak) 0.525
T Sa = (sedang)
T
0.918 0.700
0.413
Sa = (sedang) 0.280
T Sa = (keras)
0.734 T
0.525
0.688 0.220
Sa = (keras)
T
0.550 0.420 0.420

0.413 0.350 0.280


0.367 0.280
0.367 0.280 0.263
0.275 0.245 0.175
0.220 0.207
0.220 0.140
0.138 0.093
0.110 0.073

0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0
T T

Sa Wilayah 5 Sa Wilayah 6
1.200

1.100
0.840 0.960
1.105 Sa = (lunak ) 1.000 Sa = (lunak )
T 0.960 T
0.978 0.660
0.587
0.840 Sa = ( sedang ) Sa = ( sedang )
T T
0.850
0.340 0.400
Sa = (keras ) 0.660 Sa = ( keras )
T T
0.587
0.480 0.480
0.420 0.440 0.400
0.442
0.400 0.320
0.391 0.340 0.330
0.280
0.340 0.294 0.220
0.196 0.200
0.170 0.133
0.113

0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0
T T

21
SPEKTRUM RESPONS NOMINAL C (R=8.5, µ=5.3, K=1)

C Wilayah 1 C Wilayah 2
0.0735 0.0588
C= (lunak)
T
0.0318
0.0588 C= (sedang)
T
0.0530
0.0141
C= (keras)
T
0.0176
C= (lunak)
T 0.0353
0.0318
0.0080 0.0294
C= (sedang) 0.0294
T
0.0220 0.0196
0.0035 0.0212
C= (keras)
0.0176 T 0.0159
0.0141 0.0141
0.0133 0.0106
0.0080 0.0088 0.0071
0.0088 0.0059 0.0047
0.0053 0.0035 0.0040
0.0018 0.0027
0.0035 0.0012
0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0

T T

C Wilayah 3 C Wilayah 4
0.1178

0.0942
C= (lunak)
0.1030 0.0942 T
0.0864
C= (lunak) 0.0618
T C= (sedang)
T
0.1080 0.0486 0.0823
C= (sedang) 0.0329
T C= (keras)
0.0864 T
0.0810 0.0259 0.0618
C= (keras)
T
0.0471
0.0648 0.0471

0.0486 0.0412 0.0314


0.0432 0.0432 0.0329 0.0329 0.0309
0.0324 0.0288 0.0206
0.0259 0.0243
0.0259 0.0165
0.0162 0.0110
0.0130
0.0086

0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0
T T

C Wilayah 5 C Wilayah 6
0.1413

0.1295
0.1040 0.1130
0.1300 C= ( lunak ) 0.1178 C= ( lunak)
T 0.1130 T
0.1150 0.0777
C = 0.0690 ( sedang) C= ( sedang)
0.1040 T T
0.1000
0.0400 0.0471
C= (keras ) 0.0777 C= (keras )
T T
0.0690
0.0565 0.0565
0.0520 0.0518 0.0471
0.0520
0.0460 0.0471 0.0389 0.0377
0.0400 0.0347
0.0400 0.0345 0.0259
0.0230 0.0236
0.0200 0.0157
0.0133

0 0.2 0.4 0.6 0.8 1.0 2.0 3.0 0 0.2 0.4 0.6 0.8 1.0 2.0 3.0
T T

22

Você também pode gostar