CE 490 INTRODUCTION TO

EARTHQUAKE ENGINEERING

CHAPTER 6

SEISMIC ANALYSIS PROCEDURES

IN SEISMIC CODES
 Outline
▪ Current Design Approaches for Earthquake
Engineering
▪ Force-Based Design Approach
▪ Earthquake Resistant Design Principles
▪ Analysis Procedures in Seismic Codes
▪ Modal Response Spectrum Analysis
▪ Equivalent Static Lateral Force Procedure
▪ Deformation Control in Seismic Codes
▪ Drift limits
▪ Second-order effects
▪ Building separations
 Current Design Approaches for Earthquake
Engineering
Force-based approach (common for most of the codes,
reducing elastic response through the employment of R factors)

Displacement-based approach (popular for the last decade,

poten-tial for replacing classical approach, but includes more
complicated analysis methods)

Energy-based approach (not yet common in design and

analysis, but promising as it provides practical tools for the
design of structural systems with base isolation or energy
dissipation mechanisms, i.e. dampers)

The latter two approaches can be regarded in the context

of “performance-based earthquake engineering” !
 Important Note !

The “design” ground motion can only be predicted

with an acceptable order of magnitude.
Even if the motion can be predicted in a perfect
manner, it is unlikely than we can precisely predict
the response of the structure. This is due to the
rather long list of things we do not know and can not
do (aleatory), as well as uncertainties in the things
we do know and can do (epistemic).
The best we can hope for is to predict the charac-
teristics of the ground motion and the characteristics
of the response within an acceptable level of
accuracy.
 Force-Based Design Approach
Elastic analysis with reduced forces

Inelastic displacement (damage) is inevitable, but it

should be controlled

Ductile failure modes should be dominant in the

case of severe earthquake excitation

Sudden and complete collapse of the building

should be prevented

Seismic code is based on “Life Safety” criterion

under design earthquake (475 yr return period)
 Earthquake Resistant Design Principles
You should always design simple and regular
structures!
In practice, we design nearly symmetrical structu-
res in both directions such that vibration modes are
uncoupled with well-separated translational periods
and a rotational period shorter than the others
(torsionally stiff system).
Rigid floor diaphragm means uniform distribution of
lateral forces to vertical members in a typical story.
Foundations should be strong to transmit forces
safely to the ground without imposing additional
deformations to the structure above.
 Plan Irregularity
For regular buildings, there should be no plan or vertical
irregularities.

Plan irregularity (large projections in floor plan)

 Plan Irregularity

Rectangular Non-parallel axes

L-shaped

Regular with
projections Highly irregular

SIMPLE IS ALWAYS SAFE!

Plan Irregularity

Seismic joint

Rigid wall

SIMPLE IS ALWAYS SAFE!

Distribution of Vertical Members
 Vertical Irregularity
For regular buildings, there should be no plan or vertical
irregularities.
Vertical irregularity (soft story / weak story)
The seismic codes enforce continuity of lateral load resisting
system from foundation to top floor without a significant
change in stiffness, strength, mass or floor area.

Soft Story (stiffness based) Weak Story (strength based)

IDR i  A eff ,i
 2.0  0.8
IDR i +1  A eff ,i +1
 Aeff =  Aw +  Ac + 0.15  Ainf
 Vertical Irregularity
For regular buildings, there should be no plan or vertical
irregularities.
Vertical irregularity (discontinuity in vertical members and
set-backs)

L1/L2 < 1.3 (ASCE7)

L1/L2 < 1.25 (EC8)
 Vertical Irregularity

SOFT STORY / WEAK STORY

SUDDEN CHANGES IN VERTICAL STRENGTH AND RIGIDITY

Observed Damage
Observed Damage
 Analysis Procedures in Seismic Codes
The analysis procedures in seismic codes intend to
consider all conceivable uncertainties related to the building
characteristics and calculated response.
They also exploit building properties which may lead to
simplifications in analytical models and dynamic analysis
procedures.
Rigid floor diaphragm is the most prominent property of
buildings which cause enormous reduction in the number of
dynamic degrees of freedom.
The rigid floor diaphragm assumption should be verified
according to seismic code requirements. When the ratio of
void area to total floor area is larger than a limit (generally
1/3), or the slab is not stiff enough in its own plane to
transmit the lateral inertial forces, then this assumption
should not be used.
 Rigid Floor Diaphragm Assumption
Discontinuities in floor diaphragms are regarded as vertical
irregularities in seismic design codes, which should not exist
in regular structures and also violate rigid floor diaphragm
assumption.
 Analysis Procedures in Seismic Codes
Analysis procedures in seismic design codes prescribe
calculation of seismic design forces and deformations.
There are two basic linear elastic analysis procedures in
force-based design approaches in seismic design codes.
They are carried out under the design earthquake
represented by an inelastic (reduced) design spectrum
𝑆𝑎𝑅 (𝑇) as discussed before.
The first one is the modal response spectrum analysis
procedure and the second one is the equivalent static lateral
load procedure, which is derived from the first one as a
special case.
Equivalent static lateral load procedure is applicable to
simple, regular structures.
 Modal Response Spectrum Analysis
A modal superposition analysis is carried out for the
condensed system with dynamic DoFs, under the reduced
design earthquake SaR by using a sufficient number of
modes. This method is applicable to all buildings without any
restrictions.
It is based on calculating the modal forces fn as (recall Eq
5.61)
𝐿𝑛𝑖
𝑓𝑛 = 𝑚 𝜙𝑛 𝑆𝑎𝑅,𝑛𝑖 (6.1)
𝑀𝑛

where
𝐿𝑛𝑖 = 𝜙𝑛𝑇 𝑚 𝑙𝑖 ; 𝑖 = 𝑥 𝑜𝑟 𝑦 (6.2)
The force vector 𝑓𝑛 contains two lateral force components
𝑓𝑥𝑖𝑛 and 𝑓𝑦𝑖𝑛 , and one rotational moment 𝑀𝜃𝑖𝑛 at each floor i
of a 3D building.
 Modal Response Spectrum Analysis
The formulation for fn can be given as

𝑓𝑥1𝑛 𝑓𝑦1𝑛 𝑀𝑧1𝑛 𝑓𝑥2𝑛 𝑓𝑦2𝑛 𝑀𝑧2𝑛 ȁ

𝑓𝑛𝑇 = (6.3)
… … . . . 𝑓𝑥𝑁𝑛 𝑓𝑦𝑁𝑛 𝑀𝑧𝑁𝑛

Modal force components at the nth mode, acting at the mass

centre of the jth floor slab
 Modal Response Spectrum Analysis
The term Mzjn is the torsional moment developed because
of the coupling between translational and rotational DoFs.
Modal force vector 𝑓𝑛 is applied to the uncondensed
building structure with all static and dynamic DoFs activated,
and all internal modal member forces and displacements at
the nth mode are calculated from equivalent static modal
analysis.
Finally, the modal results are combined by SRSS or CQC
formulation. For example, the design moment at the top end
of the front right corner column (Mn) in the above figure is
calculated by SRSS in the following equation

𝑀 = 𝑀1 2 + 𝑀2 2 + ⋯ + 𝑀𝑁 2 (6.4)
 Modal Response Spectrum Analysis
Number of Modes to Include
Then the question is; how many modes should be used to
represent the seismic behavior of the structure? In seismic
codes, there is a lower bound for the required number of
modes (which should be checked separately in both x and y
directions):
N min N min L2 N st
 Mn =   0.90  mi
* n (6.5)
n =1 n =1 M n i =1
where
N
Ln =  mi in (6.6)
i =1
N

i =1
2
(
M n =  mi  xin + mi  yin
2
+ I oi 2in ) (6.7)

It has no theoretical basis, it is just a practical assumption.

 Earthquake Resistant Design Principles
Accidental Eccentricity
Center of mass (CM) or center of
rigidity (CR) can be shifted in any
story due to several reasons, creating
an additional eccentricity not acco-
unted for in actual structural modeling
of the system. The reasons are
Unsymmetrical cracking or yielding of vertical members
during a strong earthquake, which causes a shift in CR.
Unsymmetrical distribution of non-structural walls, infill
panels
Shift in CM due to a concentrated live load.
Shift in CM due to imperfections in construction.
In seismic codes, CM is intentionally shifted by ±5% of the
edge dimensions to account for accidental torsion.
 Equivalent Static Lateral Force Procedure
In conventional design approaches, equivalent static lateral
force procedure (ESLFP) is generally preferred rather than
spectrum analysis or time history analysis due to its simplicity.
However, it should be noted that ESLFP is more conservative
than the other methods and its use is limited.

In TEC-07 and TBSC-18, ESLFP is applicable to regular

buildings (i.e. regular mass and stiffness distribution in plan and
elevation) with
a) seismic zone (I, II, III, IV) or DTS (1,2,3,4)
b) HN<25 m or BYS ≥4, 5, 6 (negligible higher mode response)
c) hb ≤ 2.0 (negligible torsional coupling)
In American and European Standards (ASCE & EC), there
exists limitation regarding period (T) instead of building height
(HN).
 Torsional Irregularity Criterion

Torsional irregularity coefficient

 i ,max
h bi =
 i ,average
where  i ,average =
1
2

 i ,max +  i ,min 
 Equivalent Static Lateral Force Procedure
Let’s consider the effective modal forces and the resulting
modal displacements at the first mode of a building.

Then the base shear force at the 1st mode can be obtained as
𝐿21
𝑉𝑏1 = 𝑀1∗ 𝑆𝑎𝑅,1 ≡ 𝑆𝑎𝑅,1 (6.8)
𝑀1
 Equivalent Static Lateral Force Procedure
Recalling Eq 5.61 , the modal force at

the jth floor in the 1st mode (i.e. fj1) can be written as
𝐿1
𝑓𝑗1 = (𝑚𝑗 𝜙𝑗1 ) 𝑆𝑎𝑅,1 (6.9)
𝑀1

When we multiply and divide the RHS by 𝐿1 and rearrange

𝐿21 1
𝑓𝑗1 = . (𝑚𝑗 𝜙𝑗1 ) 𝑆𝑎𝑅,1 (6.10)
𝑀1 𝐿1

Substituting 𝑉𝑏1 from Eq 6.8 into Eq 6.10, we obtain

𝑚𝑗 𝜙𝑗1
𝑓𝑗1 = 𝑉𝑏1 (6.11)
𝐿1
 Equivalent Static Lateral Force Procedure
In this equation, 𝐿1 can be defined as
𝑁

𝐿1 = 𝜙1𝑇 𝑚 1 ≡ ෍ 𝑚𝑖 𝜙𝑖1 (6.12)

𝑖=1

Finally, substituting 𝐿1 from Eq 6.12 into Eq 6.11

𝑚𝑗 𝜙𝑗1 (6.13)
𝑓𝑗1 = 𝑉𝑏1
σ𝑁
𝑖=1 𝑚𝑖 𝜙𝑖1

Multiplying and dividing the RHS by the gravitational accele-

ration g gives
𝑤𝑗 𝜙𝑗1 (6.14)
𝑓𝑗1 = 𝑉𝑏1
σ𝑁
𝑖=1 𝑤𝑖 𝜙𝑖1
 Equivalent Static Lateral Force Procedure
When the 1st mode (in the direction of earthquake excitation)
is dominant on total dynamic response, then we can assume
𝑉𝑏 ≅ 𝑉𝑏1 ≡ 𝑀1∗ 𝑆𝑎𝑅,1 (6.15)

This also means that Eq 6.14 can be modified as

𝑤𝑗 𝜙𝑗1 (6.16)
𝑓𝑗 ≅ 𝑓𝑗1 = 𝑉𝑏 𝑁
σ𝑖=1 𝑤𝑖 𝜙𝑖1

The components of the first mode vector 𝜙1 in the direction of

earthquake excitation can be approximated with a linear
variation over the building height in simple buildings with
regular height wise distribution of mass and stiffness.
𝜙𝑗1 = 𝛼 𝐻𝑗 (6.17)
 Equivalent Static Lateral Force Procedure
Hj is the height of the jth floor from the base and 𝛼 is an
arbitrary constant representing the slope of linear distribution.
Substituting 𝑗1 from Eq 6.17 into Eq 6.16, we obtain
𝑤𝑗 𝐻𝑗
𝑓𝑗 = 𝑉𝑏 𝑁 (6.18)
σ𝑖=1 𝑤𝑖 𝐻𝑖
In Eq 6.18, wi is called as the seismic weight at the ith floor.
Seismic weight is the combination of dead load (gi) and live
load (qi) as
𝑤𝑖 = 𝑔𝑖 + 𝑛𝑞𝑖 (6.19)
and n is the live load reduction factor for dynamic mass.
Typical values in design codes are 0.3 for residences and
offices, 0.6 for schools, dormitories, concert halls, restau-
rants and shops.
 Equivalent Static Lateral Force Procedure
The distribution of lateral forces along the stories, given in Eq
6.18, is slightly modified in the seismic codes as
𝑤𝑗 𝐻𝑗
𝑓𝑗 = 𝑉𝑏 − ∆𝐹𝑁 (6.20)
𝑁
σ𝑖=1 𝑤𝑖 𝐻𝑖

Where FN is the additional lateral force applied at the roof

level in order to account for higher mode effects
∆𝐹𝑁 = 0.0075 𝑁 𝑉𝑏 (6.21)
In this equation, N is the total number of stories. For low-rise
rigid structures higher mode effect is less when compared to
high-rise flexible structures.
ESLFP in Seismic Codes (Summary)
𝑤𝑗 𝐻𝑗 𝑤𝑖 = 𝑔𝑖 + 𝑛𝑞𝑖
𝑓𝑗 = 𝑉𝑏 − ∆𝐹𝑁
σ𝑁
𝑖=1 𝑤𝑖 𝐻𝑖 ∆𝐹𝑁 = 0.0075 𝑁 𝑉𝑏

FN
 Estimation of the Fundamental Period
The first mode period T1 is required in ESLFP procedure to
calculate the reduced base shear force from the following
seismic code formulation
𝑊 𝑆𝑎𝑒 (𝑇1 )
𝑉𝑏 = 𝑀. 𝑆𝑎𝑅,1 ≡ (6.22)
𝑔 𝑅 (𝑇1 )
T1 can be calculated by an approximate procedure, called
Rayleigh’s method. Let Ff be a lateral force distribution
and df be the resulting lateral story displacements, obtain-
ed by static analysis. Then
1Τ2
σ𝑁 𝑚
𝑖=1 𝑖 𝑖 𝑑 2
(6.23)
𝑇1 = 2𝜋
σ𝑁𝑖=1 𝐹𝑖 𝑑𝑖

Ff can have any distribution. However using 𝑓𝑗 in Eq 6.18 for

Ff increases accuracy.
 Estimation of the Fundamental Period
There are also empirical period formulations which have
been obtained through statistical treatment of measured
building data by using non-destructive techniques or by
expert opinion. For building frames with H<40 meters, a
simpler approximate equation is employed in Eurocode 8
𝑇1 = 𝐶𝑡 ∙ 𝐻 3/4 (6.24)
where Ct=0.075 for RC buildings and Ct=0.085 for steel
buildings. In ASCE 7, hre empirical formulation is
𝑇1 = 𝐶𝑡 ∙ 𝐻 𝑥 (6.25)
where Ct=0.047, x=0.9 for RC buildings and Ct=0.072, x=0.8
for steel buildings. The simplest formulation is a rule of
thumb for regular frames with N<12
𝑇1 = 0.1 𝑁 (6.26)
 Estimation of the Fundamental Period
In TBSC-18, the first alternative is the Rayleigh method as
stated in Eq 6.23.
The second alternative is the empirical equation as stated in
Eq 6.25. The coefficients Ct and x take the following values:
for cast-in-situ and precast concrete frame buildings,
Ct=0.047, x=0.9, for steel frame buildings, Ct=0.072, x=0.8,
for other construction systems, Ct=0.049, x=0.75.
For RC shear-wall buildings, Ct can be calculated as
0.1
𝐶t = ≤ 0.07 (6.27)
𝐴𝑡
where 2
ℓ𝑤𝑗
𝐴t = ෍ 𝐴𝑤𝑗 0.2 + ≤ ෍ 𝐴𝑤𝑗 (6.28)
𝐻𝑁
𝑗 𝐽
 Deformation Control in Seismic Codes
In force-based design philosophy, buildings designed for
reduced seismic forces may undergo large inelastic
deformations, which can cause severe damage in vertical
structural elements and non-structural elements (for example;
infill wall damage in the 2011 Van earthquake)
 Deformation Control in Seismic Codes
Hence seismic codes impose limitations on interstory drift for
preventing excessive relative deformations of the stories. The
basic reason for limiting interstory drift is to prevent extensive
damage of non-structural components such as infill walls,
partition walls. This limit is generally quite high for most of the
buildings and generally does not control the design.

Interstory drift

Interstory drift ratio

(IDR)
 Deformation Control in Seismic Codes
ASCE 7 limits the interstory drift ratio to 2.5% under
earthquake forces which are “not” reduced by the response
reduction factor. Accordingly, the interstory drift deformations
obtained under the reduced design earthquake forces are re-
multiplied by a 𝐶𝑑 factor, and checked for the 2.5% limit of
interstory drift ratio at each story j.
𝐶𝑑 ∆𝑗,𝑚𝑎𝑥 ≤ 0.025 (6.27)
Eurocode 8 always has a strict drift limitation under a service-
ability earthquake (TR=95 years) and the interstory drifts are
calculated from the equal displacement rule. When converted
from the serviceability to the design earthquake, the limiting
interstory drift ratio is 0.01 for buildings with brittle partitions,
0.015 for ductile partitions and 0.02 for buildings without
partitions in contact with the frame.
 Deformation Control in Seismic Codes
For the national codes, the drift limitation in TEC-07 can be
given as follows:
𝑅∆𝑗,𝑚𝑎𝑥 ≤ 0.02 (6.28)

In the latest version of the code (TBSC-18), can be given as

follows: effective interstory drift (di) can be obtained from the
reduced interstory drift values (i) as follows:
𝑅 (6.29)
𝛿𝑖 = ∆𝑖
𝐼
In the presence of non-ductile infill walls with no structural
connection to the frame elements (only friction in betwen):
𝛿𝑖,𝑚𝑎𝑥
≤ 0.01 (6.30)
ℎ𝑖
 Deformation Control in Seismic Codes
In the presence of non-ductile infill walls being isolated from
the frame elements by proper separation:
𝛿𝑖,𝑚𝑎𝑥
≤ 0.02 (6.31)
ℎ𝑖
If Eqs 6.30-6.31 are not satisfied in a single story and/or in a
single earthquake direction, the stiffness of the structural
system should be increased by re-dimensioning the members.
 Second Order (P-Delta Effects)
Second order or P-Δ effects are generated due to the
interaction of vertical forces P with the lateral seismic
deformations Δ to produce additional overturning moments PΔ
on the vertical members of the structure.

Reduced Strength
and Stiffness

VP = V (1 −  )
K P = K (1 −  )
 Second Order (P-Delta Effects)
In the above equations, parameter  is called as the drift
sensitivity coefficient, which is simply the ratio of the second
order overturning moment to the first order overturning
moment.
𝜃 = 𝑃 ∆Τ𝑉 ℎ (6.32)

P-𝛥 effects can be indirectly and approximately compensated

in design by increasing the seismic forces, and accordingly the
lateral stiffness by a ratio 1/(1 − 𝜃).

𝜃 < 0.1 P- effects can be ignored

0.1 < 𝜃 < 0.2 Increase seismic design forces by 1/(1-)
𝜃 > 0.2 Exact second order analysis is required
𝜃 > 0.3 It is not permitted in design
 Building Seperations
Another problem is the pounding (impact) of two adjacent
buildings due to excessive deformations and insufficient
seperation. Hence a limit is required for deformation in
earthquake codes.

These two building blocks should be separated by at least a

distance d equal to the maximum value of SRSS combination
of the average displacements at the adjacent story levels.
 Building Seperations
According to EC-8 and TEC-07, the below criterion is used:

2 2
𝑑>𝛼 𝑢1,𝑎𝑣 + 𝑢2,𝑎𝑣 𝑚𝑎𝑥 (6.33)

where 𝑢𝑖 is the elastic displacement under the (reduced)

design earthquake, amplified by the R factor (equal
displacement rule). The value of 𝛼 is 0.7 if the story levels
are equal, or 1.0 if story levels are not equal in the adjacent
buildings according to EC-8.
According to ASCE 7, the minimum separation is calculated
for individual building blocks as,
𝐶𝑑 𝛿𝑥𝑒 (6.34)
𝑑>
𝐼
 Building Seperations
Here, 𝐶𝑑 is the deflection amplification factor (5.5 for special
moment resisting frames), 𝛿𝑥𝑒 is the elastic displacement at
level x under the design earthquake (reduced by R), and I is
the structure importance factor.
In TBSC-18, the following criteria are used:
1) The seperation distance will be calculated by multiplying
SRSS of maximum reduced displacements at the floor
level by Ra factor.
2) Minimum separation distance will be 30 mm up to 6 m
height and 10 mm will be added for each extra 3 m above
the height of 6 m.
 Equivalent Static Lateral Force Procedure
Example 6.1: For the 3-story shear
frame and given design spectrum,
determine the equivalent static
lateral load distribution. Let R=1 for a
consistent comparison with the
modal force vectors from spectrum
analysis. Assume h=3 meters.
𝑚 = 175 𝑡 𝑘 = 140 000 𝑘 𝑁Τ𝑚

𝑆𝑎𝑒 𝑇 = 0.4
First obtain the design base shear Vb 𝑉𝑏 = 𝑀. 𝑆𝑎𝑅,1 = 5𝑚 = 8583.75 kN
𝑅

Then calculate FN=0.0075 N Vb ∆𝐹𝑁 = 0.0075 × 3 × 8583.75 = 193.13 𝑘𝑁

𝑤𝑗 𝐻𝑗
𝑓𝑗 = 𝑉𝑏 − ∆𝐹𝑁
Next calculate fj (j=1-3) σ𝑁
𝑖=1 𝑤𝑖 𝐻𝑖
 Equivalent Static Lateral Force Procedure
350𝑔 × 3
𝑓1 = 8583.75 − 193.13 = 1864.58 𝑘𝑁
350𝑔 × 3 + 350𝑔 × 6 + 175𝑔 × 9

350𝑔 × 6
𝑓2 = 8583.75 − 193.13 = 3729.16 𝑘𝑁
350𝑔 × 3 + 350𝑔 × 6 + 175𝑔 × 9

175𝑔 × 9
𝑓3 = 8583.75 − 193.13 = 2796.87 𝑘𝑁
350𝑔 × 3 + 350𝑔 × 6 + 175𝑔 × 9

But note that for the top story, 𝑓3 + ∆𝐹𝑁 = 2796.87 + 193.13 = 2990 𝑘𝑁

2990 kN 2408 kN
Base shear obtained from ESLFP
3729 kN 3302 kN is 19% more than the one obtained
from RSA analysis.
1865 kN 1509 kN
ESLFP is more conservative than
RSA although it is more practical.
Vb=8584 kN Vb1=7219 kN

ESLFP Results RSA Results

(1st mode)