EARTHQUAKE ANALYSIS OF STRUCTURES –
DYNAMIC ANALYSIS BY MODAL SUPERPOSITION
(Cl. 7.7.5
IS 1893 : 2016 (Part 1))
Why Dynamic Analysis?
Expressions for design load calculation (cl.
7.5.3; Vb= Ah * W called Equivalent Lateral
Static Load Procedure) and load distribution with
height is based on following assumptions
Fundamental mode (first mode) dominates the response
Mass and stiffness distribution are evenly distributed with
building height
Thus, giving regular mode shape
Why Dynamic Analysis? (contd…)
In tall buildings, higher modes can be quite
significant.
In irregular buildings, mode shapes may be
quite irregular
Hence, for tall and irregular buildings, dynamic
analysis is recommended.
Note that industrial buildings may have large
spans, large heights, and considerable
irregularities:
These too will require dynamic analysis.
DIFFERENT TYPES OF DYNAMIC ANALYSIS
There are various types of Dynamic Analysis ;
namely
a) Linear Dynamic Analysis By Modal
Superposition (Cl. 7.7.4.5). Code IS 1893 calls it
Dynamic Analysis by Response Spectrum
Method (Cl. 7.8.4.2)
b) Pushover Analysis
c) Non linear Time History Analysis
In this presentation (and course), we will
concern ourselves only with Modal
Superposition Analysis
Requirement of linear Dynamic Anal. by Modal
Superposition
Cl. 7.8.1 – old code IS 1893 (Part 1) : 2002
Seismic Regular Irregular
Zone Building Buildings
II and III Height > Height >
90m 40m
IV and V Height > Height >
40m 12m
As per provisions in new code IS 1893 (Part 1) : 2016 (Cl. 7.7.1),
linear Dynamic analysis is mandatory for all buildings more than
15 m high in Seismic Zone II. Further, it is mandatory for any
Building in Zones III - V
DYNAMIC ANALYSIS BY MODAL SUPERPOSITION/
RESPONSE SPECTRUM METHOD (Cl. 7.7.5)
This analysis procedure is valid when a building can be
modeled as a lumped mass model with one degree of
freedom per floor
If the building has significant plan irregularity, it
requires three degrees of freedom per floor and the
procedure of Cl. 7.7.5 is not valid. In that case, software
modelling is advised as per Cl. 7.7.2.
Lumped Mass Model for Cl. 7.7.5
X3(t)
X2(t)
X1(t)
Value of Damping
Damping to be used
Steel buildings: 2% of critical
RC buildings: 5% of critical
For masonry buildings? Not specified.
Implies that a steel building will be designed for
about 40% higher seismic force than a similar RC
building (Table 3 of IS 1893 : 2002 gives
multiplying factor of 1.40)
Choice of damping has implications on seismic
safety. Hence, damping value and design
spectrum level go together
Number of Modes Cl. 7.7.5.2
Code requires sufficient number of modes so
that at least 90% of the total seismic mass is
excited in each of the principal directions.
Modal Mass (Cl. 7.7.5.4 (a))
The modal mass (Mk) of mode k is given by
2
n
i ik
W
M k i n1
g Wi ik
2
i 1
Where
g = Acceleration due to gravity,
Φik= Mode shape coefficient at floor i in mode k, and
Wi = Seismic weight of floor i.
Modal Participation Factors (Cl. 7.7.5.4 (b))
The modal participation factor (Pk) of mode k is
given by
W i ik
Pk n
i 1
Wi ik
2
i 1
Design Lateral Force at Each Floor in Each Mode
(Cl. 7.7.5.4 (c))
The peak lateral force (Qik) at floor I in mode k is
given by
Qik = Ak ik Pk Wi
Where
Ak = Design horizontal acceleration
(=(Z/2)*(I/R)*(Sa/g)) as per 6.4.2 using the
natural period of vibration (Tk) of mode k.
Storey Shear Forces in Each Mode
(Cl. 7.7.5.4 (d))
)
The peak shear force (Vik) acting in storey I in mode
k is given by
n
Vik Qik
i
Modal Combination Cl. 7.7.5.4e) & 7.7.5.3
This clause gives CQC method first and then
simpler method as an alternate.
CQC is a fairly sophisticated method for modal
combination. It is applicable both when the
modes are well-separated and when the modes
are closely-spaced.
Many computer programmes have CQC method
built in for modal combination.
Modal Combination Cl. 7.7.5.2 (contd…)
Response Quantity could be any response
quantity of interest:
Base shear, base moment, …
Force resultant in a member, e.g.,
Moment in a beam at a given location, Axial force in column,
etc.
Deflection at a given location
Alternate Method to CQC
Use SRSS (Square Root of Sum of Squares) if the
natural modes are not closely-spaced.
....
2
1
2
2
2
3
2
4
Use Absolute Sum for closely-spaced modes
(MAXIMUM ABSOLUTE RESPONS (ABS))
1 2 3 4 ...
To appreciate the alternative method, consider two
examples.
Example 1 on Modal Combination:
For first five modes of vibration, natural period/
natural frequency and maximum response are
given. Estimate the maximum response for the
structure.
Mode 1 2 3 4 5
Natural 0.95 0.35 0.20 0.14 0.11
Period
Natural 1.05 2.86 5.00 7.14 9.09
Frequency
Response 1100 350 230 150 120
Quantity
Example 1 on Modal Combination (contd…)
Each natural frequency differs from the higher
frequency by more than 10% of lower freq.
As per Cl. 3.1, none of the modes are closely-spaced
modes.
As per section b) in Cl. 7.7.5.3, we can use
Square Root of Sum of Squares (SRSS) method
to obtain resultant response as
(1100 ) 2 (350) 2 (230) 2 (150) 2 (120) 2 1193
Example 2 on Modal Combination
For first six modes of vibration, natural period/
natural frequency and maximum response are
given. Estimate the maximum response for the
structure.
Mode 1 2 3 4 5 6
Natural period 0.94 0.78 0.74 0.34 0.26 0.25
(sec)
Natural frequency 1.06 1.28 1.35 2.94 3.85 4.00
(Hz)
Response Quantity 850 230 190 200 90 80
Example 2 on Modal Combination (contd…)
As per Cl. 3.1, modes 2 and 3 are closed spaced since
their natural frequencies are within 10% of the lower
frequency.
Similarly, modes 5 and 6 are closely spaced.
Combined response of modes 2 and 3 as per section b)
in Cl.7.7.5.4 = 230+190=420
Combined response of modes 5 and 6 = 90 + 80 = 170
Combined response of all the modes as per section b)
(850) 2 (420) 2 (200) 2 (170) 2 984
Storey Shear Forces due to All Modes
Considered (Cl. 7.7.5.4 (e))
The peak storey shear force (Vi) in storey i due to all
modes considered is obtained by combining those
due to each mode in accordance with 7.7.5.4.
Lateral Forces at Each Storey Due to All Modes
Considered (Cl. 7.7.4.5 (f))
The design lateral forces, Froof and Fi , at roof and at
floor i (respectively)
Froof = Vroof , and
Fi = Vi - Vi+1
Lower Bound on Seismic Force (Cl. 7.7.3)
This clause requires that in case dynamic
analysis gives lower design forces, these be
scaled up to the level of forces obtained based
on empirical T.
Implies that empirical T is more reliable than T computed
by dynamic analysis
All the response quantities to be scaled up by VB /VB
FOR MORE DETAILS OF THIS CLAUSE,
SEE THE PPT ON “ DYNAMIC ANALYSIS
-IS 1893_SOLVED EXAMPLE”
Lower Bound on Seismic Force (Cl. 7.7.3)
Lower Bound on Seismic Force (Cl. 7.8.2) (contd…)
There are considerable uncertainties in modeling
a building for dynamic analysis, e.g.,
Stiffness contribution of non-structural elements
Stiffness contribution of masonry infills
Modulus of elasticity of concrete, masonry and soil
Moment of inertia of RC members
Depending on how one models a building, there
can be a large variation in natural period.
Ignoring the stiffness contribution of infill walls
itself can result in a natural period several times
higher
