CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Response to Ground Motion

Ground Motion Characteristics:

An earthquake occurs when rocks break and slip along a fault in the earth. During an
earthquake energy is released in several forms, including as movement along the fault, as
heat, and as seismic waves that radiate out from the centre of the earthquake in all
directions and cause the ground to shake, sometimes hundreds of kilometres away.
Earthquakes are caused by the slow deformation of the outer, brittle portions of tectonic
plates, the earth's outermost layer of crust and upper mantle. Due to the heating and
cooling of the rock below these plates, the resulting convection causes the adjacently
overlying plates to move, and, under great stress, deform. The rates of plate movements
range from about 2 to 12 centimetres per year. Sometimes, tremendous energy can build
up within a single, or between neighbouring plates. If the accumulated stress exceeds the
strength of the rocks making up these brittle zones, the rocks can break suddenly,
releasing the stored energy as an earthquake, which then travels through the earth in the
form of seismic waves.

Seismograph recording of the vertical (V) and horizontal (N, E) ground motion at Lillooet, BC following a
magnitude 6.6 earthquake that occurred 600 km to the west on 4 November 2, 2004. Primary (P) and
secondary (S) body waves, and Raleigh (R) surface waves indicated.

Seismic waves can be categorized into two types of wave – body waves and surface
waves. Body waves travel through the interior of the earth and consist of primary (P) and
secondary (S) waves. Both P and S waves arrive before the surface waves and are of a
higher frequency than the corresponding surface waves. P waves are compression waves
that move the ground back and forth in the direction that the wave is travelling, while S

49
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

waves can move the ground either up and down, or from side to side perpendicular to the
direction that the wave is travelling in.

Surface waves consist of Love and Raleigh waves and, although they travel much slower
than the body waves, they are primarily responsible for the damage and destruction
associated with earthquakes. Love waves move the ground from side to side producing
purely horizontal motion, while Raleigh waves roll along the ground like a wave on a
lake or the ocean, moving the ground both up and down and from side to side in the same
direction that the wave is moving in. Raleigh waves produce the largest ground motions
of the various types of seismic waves that occur in an earthquake.

Modal Analysis for Ground Motion:

So far we have considered the case of forced vibration of a structure subjected to a set of
time-varying lateral loads applied at the location of each mass, each of which corresponds
to a floor in the structure. In this case, the lowest columns in the structure are fixed to the
ground, which remains stationary, and in turn means that the motions of each mass, or
floor, are absolute, rather than relative. The question is what happens in the case where
the ground is moving laterally with velocity 𝑥̇ , as would occur in an earthquake?

In this case we can describe the motion of each mass using either absolute displacements,
or relative displacements, where the absolute displacement of mass 𝑗 is given by
𝑋 = 𝑥 + 𝑥
while the relative displacement is given by
𝑥 =𝑋 − 𝑥
The relative velocity and acceleration of mass 𝑗 with respect to the ground are then given
by the expressions
𝑥̇ = 𝑋̇ − 𝑥̇
𝑥̈ = 𝑋̈ − 𝑥̈
respectively.

Applying Newton’s second law, and assuming that the stiffness and damping matrices are
dependent on the relative displacement of mass 𝑗 with respect to the displacement of the
ground, the equation of motion for mass 𝑗 can be written as

𝑚 𝑋̈ = − 𝑘 𝑋 − 𝑥 − 𝑐 𝑋̇ − 𝑥̇ .

Substituting for the absolute acceleration in terms of the relative accelerations of the
ground and mass, and writing the displacements and velocities in terms of the relative
velocities and accelerations leads to

50
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

𝑚 𝑥̈ + 𝑥̈ =− 𝑘 𝑥 − 𝑐 𝑥̇

or, rearranging

𝑚 𝑥̈ + 𝑘 𝑥 + 𝑐 𝑥̇ = −𝑚 𝑥̈ .

Note that there is an implicit assumption that the damping is classical, and dependent on
either the mass or stiffness, which means that there is no coupling through the damping
terms. If we assume that the damping can be expressed in terms of the mass, that is
𝑐 𝑋̇ − 𝑥̇ = 2𝛼𝑚 𝑋̇ − 𝑥̇ = 2𝑚 𝛼𝑥̇
then the differential equation of motion for mass 𝑗 in terms of relative coordinates
becomes

𝑚 𝑥̈ + 2𝛼𝑚 𝑥̇ + 𝑘 𝑥 = − 𝑚 𝑥̈

This is the same equation for forced vibration that we have seen previously, with the
ground accelerations leading to a set of equivalent exciting forces such that
𝑃 (𝑡) = − 𝑚 𝑥̈ for 𝑗 = 1, 2, 3, … , 𝑛
or
− 𝑚 𝑥̈
⎧ ⎫
⋮
{𝑃(𝑡)} = − 𝑚 𝑥̈ .
⎨ ⎬
⎩ − 𝑚 𝑥̈ ⎭
Since we have assumed classical damping, we can use modal analysis to solve for the
generalized coordinates 𝑞 . Omitting the negative sign the generalized force for mode 𝑖 is
given by

𝑃 ∗ (𝑡) = {𝜙 } {𝑃(𝑡)} = 𝜙 𝑃 (𝑡) = 𝑥̈ 𝜙 𝑚 = 𝑥̈ 𝐿

where 𝐿 is the earthquake participation factor for mode 𝑖, which is equal to

𝐿 = {𝜙 } [𝑀]{1} = 𝜙 𝑚.

where {1} = [1 1 … 1] .
In matrix form the equations can be written as
[𝑀]{𝑥̈ } + [𝐶]{𝑥̇ } + [𝐾]{𝑥} = −[𝑀]{1}𝑥̈

51
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

where [𝐶] = 2𝛼[𝑀]. The matrix form of the equation can also be derived directly by
realizing that the inertial forces arise from the absolute displacements 𝑋 , while the
stiffness and damping forces are due to the relative displacements 𝑥 , which leads to
[𝑀]{𝑥̈ } + [𝐶]{𝑥̇ } + [𝐾]{𝑥} = {0}.
Substituting for 𝑋̈ in terms of the relative and ground accelerations then leads to the
equation previously derived, that is
[𝑀]{𝑥̈ } + [𝐶]{𝑥̇ } + [𝐾]{𝑥} = −[𝑀]{1}𝑥̈ .
As before, if modal analysis is used, the relative displacement of mass 𝑗 can be obtained
by summing the contributions from all modes

𝑥 (𝑡) = ∅ 𝑞 (𝑡).

The associated generalized coordinates are found by solving the equation

𝑃∗ (𝑡) 𝐿
𝑞̈ (𝑡) + 𝜂 ∗ 𝑞̇ (𝑡) + 𝜔 𝑞 (𝑡) = ∗ = 𝑥̈ (𝑡).
𝑀 𝑀∗

The exact solution to the equation will depend on the nature of the ground motion, which
could be:
 harmonic
 transient, which in the case of an earthquake would be more likely

Harmonic Ground Motion:

For harmonic ground motion, the ground displacement and acceleration are given by
expressions of the form
𝑥 (𝑡) = 𝑥 cos 𝜔𝑡
𝑥̈ (𝑡) = −𝜔 𝑥 cos 𝜔𝑡
The resulting loads on each mass 𝑗 are equal to
𝑃 (𝑡) = − 𝑚 𝑥̈ (𝑡) = 𝑚 𝑥 𝜔 cos 𝜔𝑡.
The generalized force for mode 𝑖 is then given by

𝑃 ∗ (𝑡) = 𝜙 𝑃 (𝑡) = 𝜔 𝑥 cos 𝜔𝑡 𝜙 𝑚.

The steady state response is given by the standard equation for forced vibration of a
single degree of freedom system. Setting

52
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

𝐿 = 𝜙 𝑚

the total response of mass 𝑗 is equal to

𝜔 𝐿 1
𝑥 (𝑡) = 𝑥 𝜙 cos(𝜔𝑡 − 𝜓 ).
𝜔 𝑀∗
𝜔 𝜔
1− + 4(𝜂∗ )
𝜔 𝜔

Finally, the peak resonant amplitude of mass 𝑗 for mode 𝑖 is equal to

𝐿 1
𝑥 =𝑥 𝜙 .
𝑀∗ 2𝜂∗

Transient Ground Motion:

While we considered the response of a multi-degree of freedom structure to harmonic
ground motion in the previous section, in reality earthquakes are short-lived, transient
events. In many cases we have recorded time histories of the ground accelerations
measured using seismograph networks during actual earthquakes. These have
subsequently been collected into databases that are available to engineers to use during
the design process to examine the performance of their structure to historical earthquakes
that have occurred at the structure’s location. To do this requires implementing some
form of numerical calculation of the response of the structure to a known, time-varying
input ground acceleration record. There are a number of methods available for the
numerical calculation of the response of a structure to a known, time-varying load, and in
this course we will focus on the use of Newmark’s method, although there are many other
methods that can also be used to achieve the same result.

Newmark’s Method Applied to Multi-Degree of Freedom Systems:

Newmark’s method for single degree of freedom systems can also be applied to multi-
degree of freedom systems when using modal analysis. Recall that in modal analysis we
express the displacement of each floor or mass in the system as being the sum of the
contributions to the response from the individual mode shapes, that is

{𝑥(𝑡)} = {𝜙 }𝑞 (𝑡)

where {𝑥(𝑡)} is a vector containing the time-varying displacements of the masses

forming the system, {𝜙 } is mode shape 𝑖 and 𝑞 (𝑡) is the associated modal coordinate of
mode 𝑖 that defines the time-varying contribution of that mode to the overall response of
the structure.

53
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

If we define the generalized mass, damping, stiffness, and time-varying force for each
mode, then we obtain a set of 𝑛 independent, uncoupled equations for the modal
coordinates 𝑞 (𝑡), in which each mode shape can be treated as though it were a single
degree of freedom system which must satisfy the equation
𝑀∗ 𝑞̈ (𝑡) + 𝐶 ∗ 𝑞̇ (𝑡) + 𝐾 ∗ 𝑞 (𝑡) = 𝑃 ∗ (𝑡)
where 𝑀 ∗ , 𝐶 ∗ , 𝐾 ∗ , and 𝑃∗ (𝑡) are the generalized mass, damping, stiffness, and time-
varying force for mode shape 𝑖 respectively. Rather than solving for the physical
displacements, velocities, and accelerations, we use Newmark’s method to solve for the
corresponding values in modal coordinates, that is instead of solving for 𝑥(𝑡), 𝑥̇ (𝑡), and
𝑥̈ (𝑡) we solve for 𝑞 (𝑡), 𝑞̇ (𝑡), and 𝑞̈ (𝑡). Once we have obtained 𝑞 (𝑡) for each mode
shape, we can then use these values to recover the physical displacements of the floors or
masses at each time step by summing the contributions from each mode shape.

In this case the equations required to implement Newmark’s method are

1 𝛾 1 ∗ 𝛾
∆𝑃 , = ∆𝑃∗, + 𝑀∗ + 𝐶 ∗ 𝑞̇ , + 𝑀 + ∆𝑡 − 1 𝐶 ∗ 𝑞̈ ,
𝛽∆𝑡 𝛽 2𝛽 2𝛽

= ∆𝑃∗, + 𝐴 𝑞̇ , + 𝐵 𝑞̈ ,

∆𝑃 ,
∆𝑞 , =
𝐾
𝛾 𝛾 𝛾
∆𝑞̇ , = ∆𝑞 , − 𝑞̇ , + ∆𝑡 1 − 𝑞̈ ,
𝛽∆𝑡 𝛽 2𝛽

1 1 1
∆𝑞̈ , = ∆𝑞 , − 𝑞̇ , − 𝑞̈ ,
𝛽(∆𝑡) 𝛽∆𝑡 2𝛽

where the subscript 𝑖, 𝑡 indicates the incremental modal value for mode 𝑖 between time
steps 𝑡 and 𝑡 + 1, and the constants 𝐴 , 𝐵 , and 𝑘 are calculated as
1 𝛾 1 ∗ 𝛾
𝐴 = 𝑀∗ + 𝐶 ∗ 𝐵 = 𝑀 + ∆𝑡 − 1 𝐶∗
𝛽∆𝑡 𝛽 2𝛽 2𝛽
𝛾 ∗ 1
𝐾 = 𝐾∗ + 𝐶 + 𝑀∗
𝛽∆𝑡 𝛽(∆𝑡)
respectively.

As before, the modal coordinate displacement, velocity and acceleration values at time
step 𝑡 + 1 can then calculated as
𝑞, = 𝑞 , + ∆𝑞 ,
𝑞̇ , = 𝑞̇ , + ∆𝑞̇ ,

𝑞̈ , = 𝑞̈ , + ∆𝑞̈ , .

54
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

For the case of ground motion the incremental change in force ∆𝑃∗, for mode shape 𝑖
between time steps 𝑡 and 𝑡 + 1 will be given by
∆𝑃∗, = 𝐿 ∆𝑥̈ ,

where 𝐿 is the earthquake participation factor for mode 𝑖, and ∆𝑥̈ , is the incremental
change in the ground acceleration between time steps 𝑡 and 𝑡 + 1.

55
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Example

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 4.5 m

An eigenvalue analysis of the four-storey simple shear structure of page 18 yields the
following for the natural frequencies and mode shapes of the structure.
13.0 0.620 −0.733 0.523 0.249
38.2 0.814 −0.266 −0.678 −0.816
𝜔= rad⁄s Φ =
61.0 0.933 0.415 −0.490 1.000
80.1 1.000 1.000 1.000 −0.638
The structure is subjected to a full linear time history analysis using Newmark’s method
in combination with a recorded ground motion time history from the 2010 Val-des-Bois
earthquake. The table below shows the calculated modal displacements for all four modes
for five successive time steps of the analysis.

a) Using the data provided in the table below, calculate the displacements of the first
and third floors of the structure for all five time steps, accounting for contributions
from all four modes.

Time (s) 𝑞 (m) 𝑞 (m) 𝑞 (m) 𝑞 (m)

26.310 -1.313E-03 6.538E-05 5.216E-06 9.989E-08

26.315 -1.322E-03 5.476E-05 6.751E-06 -7.273E-09
26.320 -1.318E-03 4.065E-05 8.049E-06 -3.161E-08
26.325 -1.301E-03 2.336E-05 9.049E-06 4.210E-08
26.330 -1.272E-03 3.863E-06 9.581E-06 1.817E-07

56
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

57
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Response Spectrum Concepts:

Recall that for a single degree of freedom system subjected to ground motion we can
write the following equation:
𝑚𝑥̈ + 𝑐𝑥̇ + 𝑘𝑥 = −𝑚𝑥̈ (𝑡)
or, rearranging the equation slightly
𝑥̈ + 2𝜂𝜔 𝑥̇ + 𝜔 𝑥 = −𝑥̈ (𝑡).
The latter equation shows that the response of our single degree of freedom system is a
function of only the damping ratio, 𝜂, the natural frequency, 𝜔 , and the time history of
the ground motion that the structure is subjected to. If we now subject our single degree
of freedom system to a known ground motion time history and determine the peak
displacement over the period of the record for a given natural frequency and damping
ratio, then by varying the natural frequency and damping ratio we can determine the
displacement response spectrum of our single degree of freedom system to a particular
earthquake time history. Depending on whether it is the peak displacement, velocity, or
acceleration that we are interested in, we can then define the appropriate response
spectrum for these quantities.

To determine the peak displacement values and the resulting internal forces we need only
define the displacement response spectrum. We can introduce two related spectra, the
pseudo-velocity response spectrum and the pseudo-acceleration response spectrum. If the
displacement response spectrum is denoted by 𝐷(𝑇 , 𝜂), where 𝑇 = 2𝜋⁄𝜔 is the
natural period of the system, then the corresponding pseudo-velocity response spectrum,
𝑉(𝑇 , 𝜂), is related to the displacement response spectrum by the expression
𝑉(𝑇 , 𝜂) = 𝜔 𝐷(𝑇 , 𝜂)
while the pseudo-acceleration response spectrum, 𝐴(𝑇 , 𝜂), is equal to
𝐴(𝑇 , 𝜂) = 𝜔 𝐷(𝑇 , 𝜂).
Note that neither the pseudo-velocity nor the pseudo-acceleration are equal to the
corresponding peak velocities or accelerations of the system that would be calculated
from the full time history.
The pseudo-velocity can be shown to be related to the peak value of the strain energy
stored in the system during an earthquake by the equation
𝑚𝑉
𝐸 =
2
while the pseudo-acceleration is related to the peak value of the base shear by the
equation
𝑊
𝐹 = 𝑚𝐴 = 𝐴
𝑔
where 𝑊 is the weight of the structure.

58
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Since the displacement, pseudo-velocity, and pseudo-acceleration response spectra are

related, we often plot them on a combined plot, as shown above for the smoothed
response spectra from the 1940 El Centro earthquake, where the pseudo-velocity is read
from the vertical axis, and the corresponding displacement and acceleration values are
read from the right and left trending diagonal axes respectively.

59
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Response Spectrum Concepts Applied to Modal Analysis:

Response spectrum concepts can also be applied to modal analysis, although the
calculated response is not exact in the sense that the resulting displacements, velocities
and accelerations will not be identical to the values obtained from a full time history
analysis of the structure’s response to a given earthquake ground motion time history.
The peak displacements associated with the 𝑖th mode, {𝑥 } are given by
𝐿
{𝑥 } = {𝜙 }𝐷(𝑇 , 𝜂 ∗ )
𝑀∗
1 𝐿 ∗
= ∗ {𝜙 }𝑉(𝑇 , 𝜂 )
𝜔 𝑀
1 𝐿 ∗
= ∗ {𝜙 }𝐴(𝑇 , 𝜂 )
𝜔 𝑀
when written in terms of the displacement, pseudo-velocity and pseudo-acceleration
response spectra respectively. The corresponding peak effective accelerations of the
masses or floors for the 𝑖th mode can be written as
1 𝐿 ∗
𝐿 ∗
𝑥̈ = 𝜔 {𝑥 } = 𝜔 ∗ {𝜙 }𝑉(𝑇 , 𝜂 ) = ∗ {𝜙 }𝐴(𝑇 , 𝜂 ).
𝜔 𝑀 𝑀
The corresponding peak forces acting on the individual masses or floors for mode 𝑖 can
then be calculated as
𝐿
𝑓 = {𝑚} 𝑥̈ = {𝑚} {𝜙 }𝐴(𝑇 , 𝜂∗ )
𝑀∗
with the peak base shear for the 𝑖th mode, 𝐹 , being found by summing the forces acting
on each mass for that mode, that is
𝐿
𝐹 = 𝑓 = 𝑚 𝜙 𝐴(𝑇 , 𝜂∗ )
𝑀∗

𝐿
= 𝐴(𝑇 , 𝜂∗ ) 𝑚𝜙
𝑀∗

𝐿
= 𝐴(𝑇 , 𝜂∗ ) .
𝑀∗
To combine the responses from the individual modes to find the peak total response with
contributions from several or all of the modes is not a trivial task. This is because the
peak response associated with a particular mode will, in general, occur at a different time
from the times at which the peak responses of the other modes will occur. This means
that simply summing the peak responses, although it provides an upper bound to the peak
value of the total response, will lead to a value that is usually too conservative for
structural design applications. For structures with well-separated natural frequencies, the

60
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

square-root-of-the-sum (SRSS) rule for modal combinations can be used to provide a

good estimate of the peak total response. This states that the peak total response for mass
or floor 𝑗 is given by the square root of the sum of the squared responses for the
individual mode shapes, that is

𝑥, ≅ 𝑥 +𝑥 +𝑥 +⋯= 𝑥 .

For a more general rule, that overcomes the limitations of the SRSS rule, the complete
quadratic combination (CQC) rule for modal combinations can be used. This states that

𝑥, ≅ 𝜌 𝑥 𝑥

where 𝜌 is the correlation coefficient for modes 𝑘 and i. Since the value of 𝜌 lies
between 0 and 1, and for 𝑘 = 𝑖 it is equal to 1, the equation for the CQC rule can be
rewritten as

𝑥, ≅ 𝑥 + 𝜌 𝑥 𝑥 for (𝑘 ≠ 𝑖).

The first term is identical to the SRSS combination rule, while the double summation
term can be either positive or negative. This means that the estimate for the peak total
response of mass 𝑗 from applying the CQC rule can be larger or smaller than the estimate
obtained by applying the SRSS rule. Both the SRSS and CQC rules can be applied in a
similar manner to other quantities of interest, such as the total force acting on a mass, the
total base shear or the total base overturning moment.

61
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Example:

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 4.5 m

An eigenvalue analysis of the four-storey simple shear structure of page 18 yields the
following for the natural frequencies and mode shapes of the structure.
13.0 0.620 −0.733 0.523 0.249
38.2 0.814 −0.266 −0.678 −0.816
𝜔= rad⁄s Φ=
61.0 0.933 0.415 −0.490 1.000
80.1 1.000 1.000 1.000 −0.638
while the masses are
𝑀 7500
𝑀 12500
= kg.
𝑀 15000
𝑀 20000
The modal damping ratio for all modes is such that 𝜂 ∗ = 0.02. The generalized masses,
𝑀 ∗ , have previously been calculated as
𝑀∗ 26781
𝑀∗ 22867
= kg.
𝑀∗ 21460
𝑀∗ 36008
a) Calculate the earthquake force participation factor 𝐿 for all modes.
b) Calculate the values of the pseudo-velocity and pseudo-acceleration for each mode
using the smoothed El Centro response spectrum on page 59 of the notes
c) The peak displacement of the highest floor, the peak base shear, and the peak base
overturning moment for the first mode only.
d) The peak total displacement of the highest floor, the maximum interstorey drift at the
highest floor and the peak total base shear using the SRSS rule.

62
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

63
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

64
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

National Building Code of Canada (NBCC) Approaches to

Ground Motion:

While seismic activity in Canada is considered to be low to moderate when compared to

other regions of the world such as California, Japan or Turkey, over 1500 earthquakes are
recorded on average each year in Canada. Of these 1500 earthquakes only a few dozen
are large enough to be felt, however large and damaging earthquakes have occurred
within Canada in the recent past. The most seismically active region in Canada is
Western Canada with over 1000 earthquakes recorded annually. Much of this activity
occurs along the active plate boundaries off the coast of British Columbia, and in
particular along the boundaries of the Explorer and Juan de Fuca plates and in the
Cascadia Subduction Zone where the Juan de Fuca plate dips beneath the easterly
neighbouring North American Plate. North of the Explorer and Juan de Fuca plates lies
the Queen Charlotte fault which marks the boundary between the Pacific and North
American plates and is the site of the largest earthquake ever recorded in Canada, a
magnitude 8.1 event that occurred on 22 August, 1949.

Although Eastern Canada is located in a stable continental region of the North American
plate approximately 450 earthquakes occur annually in this region. Most of these
earthquakes are too small to be felt, but it should be noted that two of the top ten
earthquakes in Canada have been recorded in this region, a magnitude 7.0 event

65
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

occurring on 5 February, 1663, at Charlevoix, Quebec, and a magnitude 7.2 event being
recorded on 11 November, 1929, on the Grand Banks south of Newfoundland. The
causes for these earthquakes are not well understood as there are no plate boundaries in
this region because of its location within the stable interior of the North American plate.
Instead it is thought that this activity is related to regional stress fields with the
earthquakes concentrated in regions of crustal weakness.
The default procedure for the analysis of earthquake loads on a structure in the 2020
NBCC is the dynamic analysis procedure. This allows the designer to use either suitably
modified ground motion time histories from earthquakes previously recorded at the
location of the structure to calculate the full time history of the response of the structure,
or the design spectral acceleration spectrum specified for the location in combination
with modal analysis to determine the peak total values required for design of the
structure. The equivalent static force procedure can be used for structures meeting one of
three conditions, but all other structures must be designed using the dynamic analysis
procedure.

Design Spectral Acceleration Response Spectrum in the NBCC:

In considering the response spectrum method so far we have considered the response of a
structure to a response spectrum derived from a single earthquake ground motion time
history at a specific site. For design purposes, we need to define a response spectrum that
is representative of all of the ground motions to have been recorded at a particular site
during past earthquakes. This generally requires a statistical analysis of the response
spectra for an ensemble of ground motion time histories recorded at the site, before
setting a suitable level of risk to define the design response spectrum. In the NBCC
ground motion probability values are given in terms of the likelihood of a given
horizontal acceleration or velocity being exceeded during a particular period of time. The
probability used in the 2020 NBCC is 0.000404 per annum, equivalent to a 2-per-cent
probability of exceedance over 50 years. This means that over any 50-year period there is
a 2-per-cent chance of an earthquake causing ground motion greater than the given
expected value.

The seismic hazard at a given site is determined from numerous factors. Canada has been
divided into earthquake source regions based on past earthquake activity and tectonic
structure. The relationship between earthquake magnitude and the average rate of
occurrence for each region is weighed, along with variations in the attenuation of ground
motion with distance. In calculating the seismic hazard at a particular site all earthquake
source regions within a relevant distance of the site are considered. The seismic hazard is
then described by the resulting spectral acceleration values at periods of 0.2, 0.5, 1.0, 2.0,
5.0 and 10.0 seconds, with all spectral accelerations expressed as a fraction of gravity.
The six spectral acceleration values then allow the construction of what are called
uniform hazard spectra (UHS) for any location in Canada. In calculating the seismic
hazard a structure with damping that is 5% of critical is assumed. For structures with
damping values that are larger or smaller than this value, the spectral acceleration spectra
are used as in with no modification for the difference in damping values. This is justified

66
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

on the grounds that the difference in response due to the difference in damping values is
within the uncertainty of the code approach when used to determine peak total base shear
and deflection values.

Spectral acceleration values for a period of 0.2 seconds at a probability of 2%/50 years for firm ground
conditions (NBCC site class C). Spectral acceleration is contoured in g.

67
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Uniform hazard spectra from the 2010 NBCC for four Canadian cities for 5% damping, with acceleration
values given in g.

The design spectral acceleration response spectrum depends not only on the earthquake
history of a particular location, but also on the underlying ground conditions at the site
that the structure is to be built. The preferred method to assess the site designation is to
use the average shear wave velocity 𝑉 calculated from in situ measurements of the
shear wave velocity. The site designation is then 𝑋 where 𝑉 is the value of 𝑉 , unless
the ground profile matches one of the profiles described in Table 4.1.8.4.-A, in which
case the site designation is determined in accordance with the table.

If 𝑉 calculated from in situ measurements is not available, then the site designation is
𝑋 where 𝑆 is the site class determined from Table 4.1.8.4.-B, using either the energy-

68
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

corrected average standard penetration resistance, 𝑁 , or the average undrained shear

strength 𝑆̅ .

The design spectral response acceleration spectrum can then be calculated using Table
4.1.8.4.-C.

69
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

For the purposes of a linear dynamic analysis using the modal spectrum response method
it should be noted that the uniform hazard spectrum produced using the code approach is
not a true spectral acceleration response spectrum. Rather it is a slightly conservative
representation of the true acceleration response spectrum although the degree of
conservatism is insignificant for practical design purposes.

NBCC Equivalent Static Force Procedure:

For earthquake loads the 2020 NBCC allows a regular structure to be analyzed using the
equivalent static force procedure provided that the height of the structure is less than
60 m and the fundamental lateral period is less than 2 seconds. Certain types of irregular
structures may also be analyzed using this procedure provided that they are less than
20 m in height and have a fundamental lateral period of less than 0.5 seconds. In all other
cases the dynamic analysis procedure must be used, unless the product of 𝐼 𝐹 𝑆 (0.2) is
less than 0.35 in which case the equivalent static force procedure can again be used.
Examples of irregularities in structures include:
1) large concentrated masses or abrupt changes in mass, stiffness or strength from
storey to storey.
2) discontinuous shear walls.
3) large eccentricities in which the centre of mass and centre of twist (resistance) of
any floor in the building are widely separated.
4) low torsional stiffness.
5) vertical or horizontal geometric irregularities.
6) variations in column stiffness which attract forces to the stiffer columns.
7) diaphragm discontinuities.
8) non-parallel lateral force resisting systems.
Table 4.1.8.6 in Part 4.1 of the 2020 NBCC contains a complete description of the
irregularity types and definitions that would result in a structure being classified as an
irregular structure for design purposes under the provisions of the NBCC.

Minimum Lateral Earthquake Force:

The equivalent static force procedure in the 2020 NBCC uses the following expression to
calculate the minimum lateral earthquake force that the structure should be designed for:
𝑆(𝑇 )𝐼 𝑊𝑀
𝑉=
𝑅 𝑅
where
𝑉 = minimum lateral earthquake force
𝑆(𝑇 ) = design spectral response acceleration for the period of the first mode of

70
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

vibration
𝐼 = earthquake importance factor
𝑊 = weight of building
𝑀 = higher mode factor
𝑅 , 𝑅 = ductility and overstrength force modification factors
For walls, coupled-walls, and wall-frame systems the minimum lateral earthquake force
shall not be less than
𝑆(4.0)𝐼 𝑊𝑀
𝑉=
𝑅 𝑅
while for moment resisting frames, braced frames and other systems it shall not be less
than
𝑆(2.0)𝐼 𝑊𝑀
𝑉= .
𝑅 𝑅
For buildings located on a site that is not classed as Site Class 𝑋 and with a seismic
force resisting system (SFRS) with an 𝑅 value greater than or equal to 1.5, it need not be
greater than the larger of
2 𝑆(0.2)𝐼 𝑊 𝑆(0.5)𝐼 𝑊
𝑉= or 𝑉= .
3 𝑅 𝑅 𝑅 𝑅

Design Spectral Response Acceleration, 𝑆(𝑇 ):

To determine the design spectral response acceleration the fundamental lateral period of
the building in the direction being considered must first be calculated.

For steel moment-resisting frames this can be calculated as:

⁄
𝑇 = 0.085(ℎ ) ,
while for concrete moment frames it can be calculated as
⁄
𝑇 = 0.075(ℎ ) ,
and for other types of moment frame as
𝑇 = 0.1𝑁,
where ℎ is the height of the highest floor in the structure, and 𝑁 is the number of floors
in the structure respectively.

For braced frames it can be calculated as

𝑇 = 0.025ℎ
while for shear walls and other structures it can be calculated as
⁄
𝑇 = 0.05(ℎ ) .

71
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

These values can also be calculated using established structural models that meet the
requirements of clause 4.1.8.3(8). They must however be checked against the
requirements of clause 4.1.8.11(3)(d) to ensure the calculated values are within the
specified limits. This requires that for moment frames the calculated period does not
exceed 1.5 times the value determined using the appropriate expression given on the
previous page, while for braced frames and shear walls the calculated period must not
exceed 2.0 times the value determined using the appropriate expression given above. For
all other structures the calculated period must not exceed the value calculated using the
expression 𝑇 = 0.05(ℎ ) ⁄ . If the period calculated using an established structural
model exceeds these limits, then the limiting value is to be used as the fundamental
natural period, rather than the calculated value.

The underlying ground conditions at the site must also be established so that the site can
be classified in terms of its seismic response using either of the two approaches outlined
in the previous section. The design spectral response acceleration for the fundamental
lateral period, 𝑇 , of the structure for the direction being considered is then calculated
using Table 4.1.8.4.-C, using log-log or linear interpolation as appropriate to calculate
intermediate values between those given in the table.

Importance Factor, 𝐼 :

The value for the importance factor depends on the importance category that the building
has been assigned to, as shown in the table below.

Importance Importance Factor 𝑰𝑬

Building Use and Occupancy
Category ULS SLS
Low-human occupancy buildings
where collapse is not likely to
Low 0.8 -
cause injury, or other serious
consequences
All buildings except those listed in
Normal 1.0 -
other importance categories
Buildings likely to be used as
High post-disaster shelters, including 1.3 -
schools and community centres
Post-disaster buildings, including
hospitals, telephone exchanges,
power stations, sewage and water
treatment facilities, transportation
Post-disaster 1.5 -
control centres, emergency
response facilities including fire,
police and rescue stations, and
communication facilities

72
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Building Weight, 𝑊:

The building weight includes:

 the dead weight of the structure,
 25% of the design snow load,
 60% of any storage load, and
 the full contents of any tanks in the structure.

Higher Mode Factor, 𝑀 :

The higher mode factor is evaluated using Table 4.1.8.11 in Part 4.1 of the 2020 NBCC,
and depends on both the fundamental lateral period and the ratio of 𝑆 (0.2)⁄𝑆 (5.0) for
the location being considered, with linear interpolation being used to calculate the value
of 𝑀 for intermediate values of 𝑇 as appropriate.

73
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Ductility and Overstrength Force Modification Factors, 𝑅 and 𝑅 :

The ductility and overstrength force modification factors are evaluated using Table
4.1.8.9 in Part 4.1 of the 2020 NBCC, and depend on the type of seismic force resisting
system being used in the structure.

74
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Distribution of Lateral Seismic Forces Over Building Height:

The distribution of the lateral seismic forces over the building height is similar to the
deflected shape for the fundamental mode of vibration, which in the first instance is
assumed to be linear. The lateral seismic force at any level 𝑥 is then given by

𝐹 = (𝑉 − 𝐹 ) 𝑊 ℎ 𝑊ℎ

where
𝐹 = lateral earthquake force at level 𝑥
𝑊 = weight of floor at level 𝑥
ℎ = weight of floor at level 𝑥
𝑊 = weight of floor at level 𝑖
ℎ = weight of floor at level 𝑖

For buildings with a fundamental period greater than 0.7 seconds the higher modes of
vibration contribute significantly to the deflected shape of the structure. This effect is
accounted for by applying a concentrated load 𝐹 at the top of the building such that
𝐹 = 0.07𝑇 𝑉, 𝐹 ≤ 0.25𝑉.
If the fundamental period is less than or equal to 0.7 seconds, 𝐹 is simply equal to zero.

Finally, the total shear force for any storey 𝑥 is the sum of the lateral forces acting above
that storey.

Overturning Moments:

The overturning moment at any level 𝑥 is given by an expression of the form

𝑀 =𝐽 𝐹 (ℎ − ℎ )

where
𝐽 = 1.0 for ℎ ≥ 0.6ℎ , and
𝐽 = 𝐽 + (1 − 𝐽)(ℎ ⁄0.6ℎ ) for ℎ < 0.6ℎ .
The base overturning moment reduction factor 𝐽 that appears in the above equations is
evaluated using Table 4.1.8.11 in Part 4.1 of the 2020 NBCC.

75
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Torsion Effects:
The 2020 NBCC requires that torsion effects be considered using the worst case of
𝑇 = 𝐹 (𝑒 + 0.10𝐷 )
𝑇 = 𝐹 (𝑒 − 0.10𝐷 )
where 𝑒 is the distance measured perpendicular to the direction of earthquake loading
between the centre of mass and the centre of rigidity at the level being considered, and
𝐷 is the plan dimension of the building at level 𝑥 perpendicular to the direction of the
seismic loading being considered. The resulting torsional load then leads to additional
forces acting on individual components of the seismic force resisting system that increase
with increasing distance from the centre of rigidity of the level being considered. These
additional forces arising from torsion effects are then combined with the shear forces
acting at that level to give the required design loads at that level.

76
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Example:

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 4.5 m

An eigenvalue analysis of the four-storey simple shear structure of page 18 yields the
following for the natural frequencies and mode shapes of the structure.
13.0 0.620 −0.733 0.523 0.249
38.2 0.814 −0.266 −0.678 −0.816
𝜔= rad⁄s Φ=
61.0 0.933 0.415 −0.490 1.000
80.1 1.000 1.000 1.000 −0.638
while the masses are
𝑀 7500
𝑀 12500
= kg.
𝑀 15000
𝑀 20000
Assume that the building is to be used as an office building and is located in Orleans,
Ontario. Geotechnical site investigations show that the average standard penetration
resistance in the top 30 m of the ground, 𝑁 , in blows per 0.3 m is greater than 60. For
Orleans, Ontario, the following seismic hazard information can be obtained from the
NBCC:

Site
𝑆 (0.2) 𝑆 (0.5) 𝑆 (1.0) 𝑆 (2.0) 𝑆 (5.0) 𝑆 (10.0)
Designation
XE 0.468 0.415 0.237 0.109 0.0283 0.00848
XD 0.433 0.356 0.199 0.091 0.0234 0.0072
XC 0.409 0.239 0.121 0.0542 0.0137 0.00452
XB 0.343 0.153 0.072 0.0314 0.00779 0.00278
XA 0.242 0.122 0.0603 0.0271 0.00706 0.00265

77
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

The roof snow load has been calculated to be 348 kN using the snow loading provisions
of the NBCC.
The seismic force resisting system is composed of moderately ductile steel moment-
resisting frames with ductility and overstrength force modification factors such that 𝑅 =
3.5, and 𝑅 = 1.5.
a) Calculate the minimum lateral earthquake force the structure should be designed for
using the equivalent static force procedure of the NBCC.

78
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

79
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

80
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

81
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

NBCC Dynamic Analysis Procedure:

The 2020 NBCC dynamic analysis procedure allows the designer to perform a linear
dynamic analysis of the structure using modal analysis in combination with the specified
design spectral acceleration spectrum at the location the structure is being built.
Alternatively, suitably modified ground motion time histories from earthquakes
previously recorded at the location of the structure can be used to calculate the full time
history of the response of the structure.

When using earthquake ground motion time histories to analyze the response of a
structure in the time domain, one requirement of the NBCC is that the time histories are
compatible with the response spectrum for the site constructed from the design spectral
acceleration values. In this case compatible means that the response spectrum derived
from the ground motion time histories equals or exceeds the design spectral response
spectrum over the range of the periods of the modes contributing to the response of the
structure. This usually involves scaling recorded ground motion time histories made in
the vicinity of the site location so that the spectral acceleration associated with the period
of the first mode of vibration matches the corresponding value from the design spectral
response acceleration spectrum.

In general one ground motion time history will not be able to represent the range of
possible responses at a specific site because of the inherent variability in earthquake
ground motions, even at a specific site. This means that results from multiple records
must be used so that a realistic description of the variability in the response of the
structure can be obtained. The recommendation is that a minimum of seven correctly-
scaled individual ground motion time histories be used.

Elastic versus Inelastic Response:

Both linear dynamic analysis methods considered in the 2020 NBCC assume that the
response of the structure is elastic, and the resulting load is referred to as the elastic base
shear, 𝑉 . For structures located on sites that are classified as anything other than site
class F, with a seismic force resisting system with a ductility factor 𝑅 that is greater than
or equal to 1.5, the elastic base shear obtained from a linear dynamic analysis can first be
multiplied by a factor equal to the larger of
2 𝑆(0.2) 𝑆(0.5)
≤ 1.0 or ≤ 1.0
3 𝑆(𝑇 ) 𝑆(𝑇 )
to obtain the corresponding design elastic base shear, 𝑉 .

Allowing the structure to permanently deform by moving into the inelastic region lets the
design elastic base shear to be reduced because the deformation of the structure absorbs a
significant proportion of the energy in the earthquake. When using the equivalent static
force procedure, the minimum lateral earthquake force is calculated as
𝑆(𝑇 )𝐼 𝑊𝑀
𝑉= .
𝑅 𝑅

82
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Of particular importance in this equation are the importance factor, 𝐼 , which is based on
the intended use of the structure, and the ductility and overstrength force modification
factors, 𝑅 and 𝑅 , which account for the inelastic behaviour of the structural system
being used to resist the lateral loads imposed on the structure.

In order to obtain the design inelastic base shear, 𝑉 , from a linear dynamic analysis of
the structure, the design elastic base shear obtained from the analysis should be divided
by the product of the ductility and overstrength force modification factors, 𝑅 𝑅 , before
being multiplied by the appropriate important factor, 𝐼 . This yields a value for the design
base shear for the structure that is comparable with the value that would be obtained from
the equivalent static force procedure.

Note that the period of the fundamental mode of vibration used to calculate the base
elastic shear can be obtained from a structural model as for the equivalent static
procedure, but that this value is still subject to the same constraints as before. If the value
from the structural model exceeds the appropriate upper limit, then the upper limit value
is to be used rather for the period of the fundamental mode of vibration, rather than the
value from the structural model. This also means that in this case the effect of the higher
order modes cannot be accounted for, as they will also be higher than allowed but no
guidance is provided to estimate their values.

In general the value for the design base shear obtained from a linear dynamic analysis of
the structure will be a more accurate representation of the behaviour of the structure than
that obtained from the equivalent static force procedure. One issue, however, is that the
stiffness of the structure is usually underestimated because of a failure to account for the
contributions to the stiffness from non-structural elements. For this reason the maximum
allowable reduction in the design base shear value obtained from a linear dynamic
analysis is such that the resulting value is never less than 80% of the corresponding value
obtained from the equivalent static force procedure.

Finally, in calculating the check value for the design inelastic base shear using the
equivalent static force procedure, if the calculated period for the first mode of vibration is
obtained from an established structural model and exceeds the specified limits given in
clause 4.1.8.11(3)(d) of the 2020 NBCC, then the appropriate upper limit for the value of
the fundamental period should be used when determining the value of 𝑉 to be used as the
check value.

To obtain the values of other parameters such as the deflections, interstorey drifts
between floors, and the shear forces on individual floors of the structure, the values
obtained from a linear dynamic analysis should be scaled by the ratio of the design
inelastic base shear to the elastic base shear, that is 𝑉 ⁄𝑉 .

83
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Example:

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 3.0 m

ℎ = 4.5 m

An eigenvalue analysis of the four-storey simple shear structure of page 18 yields the
following for the natural frequencies and mode shapes of the structure.
13.0 0.620 −0.733 0.523 0.249
38.2 0.814 −0.266 −0.678 −0.816
𝜔= rad⁄s Φ=
61.0 0.933 0.415 −0.490 1.000
80.1 1.000 1.000 1.000 −0.638
while the masses are
𝑀 7500
𝑀 12500
= kg.
𝑀 15000
𝑀 20000
The modal damping ratio for all modes is such that 𝜂 ∗ = 0.02. The generalized masses,
𝑀 ∗ , and earthquake participation factors 𝐿 , have previously been calculated as
𝑀∗ 26781 𝐿 455
𝑀∗ 22867 𝐿 1665
= kg and = kg
𝑀∗ 21460 𝐿 −5963
𝑀∗ 36008 𝐿 43773
respectively.
a) Calculate the design spectral acceleration response spectrum using the provisions of
the NBCC, assuming that the building is located in Orleans, Ontario and is to be used
as an office building. Geotechnical site investigations show that the average standard
penetration resistance in the top 30 m of the ground, 𝑁 , in blows per 0.3 m is
greater than 60. For Orleans, Ontario, the following seismic hazard information can
be obtained from the NBCC:

84
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

Site
𝑆 (0.2) 𝑆 (0.5) 𝑆 (1.0) 𝑆 (2.0) 𝑆 (5.0) 𝑆 (10.0)
Designation
XE 0.468 0.415 0.237 0.109 0.0283 0.00848
XD 0.433 0.356 0.199 0.091 0.0234 0.0072
XC 0.409 0.239 0.121 0.0542 0.0137 0.00452
XB 0.343 0.153 0.072 0.0314 0.00779 0.00278
XA 0.242 0.122 0.0603 0.0271 0.00706 0.00265

b) Using the calculated design spectral acceleration response spectrum, calculate the
design elastic base shear, using modal analysis and the SRSS rule.
c) If the building is to be used as an office building, and the seismic force resisting
system is composed of moderately ductile steel moment-resisting frames with
ductility and overstrength force modification factors such that 𝑅 = 3.5, and 𝑅 =
1.5, calculate the design inelastic base shear.

85
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

86
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

87
CEE 4491A Structural Dynamics II Ground Motion and MDOF Systems

88