Response of

structures
Response spectrum

• A response spectrum is simply a plot of the peak

or steady-state response (displacement, velocity or
acceleration) of a series of oscillators of varying
natural frequency, that are forced into motion by the
same base vibration or shock.
• One such use is in assessing the peak response of
buildings to earthquakes.
• If the input used in calculating a response spectrum is
steady-state periodic, then the steady-state result is
recorded.
• Damping must be present, or else the response will
be infinite.
A plot of the peak acceleration for the mixed vertical oscillators
Typical elastic response spectra
Elastic response spectra(contd…)

The spectra shown are for elastic response to an

earthquake. That is, the structures used to develop the
curves moved and swayed during the earthquake, but
there was no yielding. For that reason, the curves are
known as elastic response spectra.
 Idealized response spectra
The response spectra derived from the behaviour of
one SDOF system in one particular earthquake are
usually quite jagged. It is not possible to use such a
historical record for design since it is unlikely that an
earthquake matching the original earthquake in duration,
magnitude or time history will reoccur.
Even if the design earthquake was completely
specified, the significant variation in spectral values over
small period ranges would require an unreasonable
accuracy in the determination of the building period.

To get around these problems, a smoothed average

design response spectrum based on the envelopes of
performance of several earthquakes is developed.
Idealized response spectra(contd…)

Average elastic design response spectra

Response spectra for other earthquakes

In order to apply the average design response spectra

to other earthquakes, they are simply scaled upward or
downward for larger and smaller earthquakes,
respectively.
Ductility
The expected magnitude of seismic loads and the
nature of building codes makes it necessary to accept
some yielding during large earthquakes. The design
provisions in modern seismic codes could not create a
purely elastic response during a large earthquake; in
any case, building a structure with such a response
would not be economical.
Displacement ductility (or just ductility) is the
capability of a structural member or building to distort
and yield without collapsing.
During an earthquake, a ductile structure can
dissipate large amounts of seismic energy even after
local yielding of connections, joints, and other
members has begun.
Ductility factor
The actual ductility of a joint or structural member is
specified by its ductility factor, . There are a number of
definitions of the ductility factor, all of which represent
the ratio of some property at failure(i.e., fracture) to that
same property at yielding. For example, the ductility
factor may be specified in terms of energy absorption.

U fracture

U yield

In addition to the definition based on the ratio of

energies, there are definitions of the ductility factor
based on the ratios of linear strain and angular
strain(rotation).
Ductility factor(contd…)
The minimum ductility of building structures with good
connections and good redundancy that are designed to
modern seismic codes seems to be about 2.5.

Desirable levels vary, although it is best to have large

values of the ductility factor-4 to 6 for concrete frames
and 6 to 8 for steel frames.
Strain energy and ductility factor
The area under the stress-strain curve represents the
strain energy absorbed U.
The maximum energy that can be absorbed without
yielding (i.e, the area under the curve up to the yield
point) is known as the modulus of resilience, UR.
The maximum energy that can be absorbed without
failure is the modulus of toughness(rupture), UT.
One definition of the ductility factor, , can be
calculated from the ratio of these two quantities

UT

UR
 Hysteresis
Hysteresis(hysteretic) damping is the dissipation of
part of the energy input when a structure is subjected
to load reversals in the inelastic range. Such
dissipation occurs in the structure itself as well as in
the soil around the foundation and, therefore, depends
on the nature of the building, foundation and soil.

The energy lost per

cycle, UH, is the area
Within the hysteresis
loop. Hysteresis Losses
are unaffected by the
Velocity of the structure.
 Hysteresis(contd…)

hysteresis damping in seismic studies is sometimes

accounted for by defining an equivalent internal
viscous damping. Such an approximation works
reasonably well in some cases, but the validity
deteriorates as the deflection increases.
Inelastic response spectra
The total seismic energy;U, received by a building
structure in an earthquake is stored or dissipated in
four primary ways
1. Some of the energy is stored as elastic strain
energy(UE)
2. Some is converted into kinetic energy (Uk)
3. Some is dissipated as hysteretic or plastic losses
(UH)
4. Some is lost due to frictional and damping effects
(Ux)
In the simplest models, the sum of these four terms
equals the total energy input.
 Inelastic response spectra(contd…)

Particularly when the building is stressed in the elastic

region, input energy is dissipated relatively slowly,
primarily because of internal friction(i.e, damping) of
the structure converting the kinetic energy into heat.
However, it takes much more energy to plastically
deform parts of the structure.
 Inelastic response spectra(contd…)

Thus, each yielding connection, every broken column,

and every sheared pin dissipates a finite amount of
kinetic energy. Therefore, a building’s amplitude of
oscillation and number of oscillation cycles decease as
major portion of the building yield.
 Inelastic response spectra(contd…)

The inelastic design response spectra (IDRS) show

what the acceleration will be when some of the seismic
energy is removed inelastically.

It is appropriate to consider the inelastic effects when

the response of a building to a major earthquake is
being determined. The inelastic response spectra are
usually derived from the elastic response spectra.
 Inelastic response spectra(contd…)
There are several well-known methods of obtaining
the inelastic response spectra from the elastic
response spectra, but few of them are suitable for
manual calculations
Perhaps the quickest and easiest, not necessarily the
most rigorous, method is simply to scale the elastic
curves downward by some function of the ductility
factor.

Sa ,elastic
Sa ,inelastic 
factor
 Inelastic response spectra(contd…)

In converting an elastic response spectrum to an

inelastic response, the ductility factor,  used to
calculate the reduction factor may be known as the
structure deflection ductility factor or design ductility
factor, . It is the ratio of the deflection at ultimate
collapse to the deflection at first yield, measured at the
roof of the structure.
 Inelastic response spectra(contd…)

At high periods (i.e, low frequencies), energy

absorption effects dominate, and a ductility factor
based on energy (rather than strain) u, is more
appropriate for use in determining the inelastic
response spectrum.
 Drift
Drift (also known as storey drift) is the deflection of
one floor relative to the floor below. There are two main
reasons to control drift.
1. Excessive drift in upper stories has strong adverse
psychological and physical effects on occupants.
2. It is difficult to ensure structural and architectural
integrity with large amounts of drift.
Architectural failures are such nonstructural damage
as failure of partitions, windows, and hung ceilings. In
low-rise construction, damage to stairwells and
elevators shafts can also be considered nonstrcutural.
However, in high-rise construction, stairwells and
elevator shafts usually constitute the most critical
structural elements in the structural core.
 Drift(contd…)
Excessive drift can be accompanied by large
secondary bending moments and inelastic behaviour.
In a severe earthquake in which yielding is
experienced, a modern high rise building can be
expected to experience a drift of approximately 1.5% of
its total height at the uppermost storey
The UBC limits the drift under the code-specified
design lateral forces to much less than this-
approximately 0.50% of the height. Under larger
forces, the drift will be larger.
 Drift(contd…)
There are three components of drift:
1. Column bending from lateral and eccentric loads
2. Joint rotation due to transverse loads
3. Frame bending from axial loads known as chord
drift.
P- effect

The column members in a building are loaded in

compression by the vertical live and dead loads.
Normally, these loads are concentric with the base of
the building. When the building is acted upon by a
lateral (horizontal) seismic load, however, the vertical
loads are eccentric with respect to the base.
P- effect
The overturning moment adds an eccentric bending
stress to the columns. The magnitude of the overturning
moment is P, where P is a function of the building
weight and  is the drift. The additional column stress is
referred to as the P-  effect.
 P- effect(contd…)
If the overturning moment increases faster than the
restoring moment from the frame stiffness, the frame is
unstable. Since the vertical load is constant (i.e. is not
transient like the seismic load that causes the initial
eccentricity), the column members will eventually fail
and the frame will buckle. Protection against instability
failures is provided by wall X-bracing and thick shear
walls.
 Torsional shear stress
A building’s center of mass (on plan view) is a point
through which the base shear (i.e., the total lateral
seismic force) can be assumed to act. This base shear
is resisted by the vertical members at the ground level.
Each such member may have a different rigidity and
thus provides a different lateral resisting force in the
opposite direction of the base shear. The building’s
center of rigidity is a point through which the resultant
of all the resisting forces acts.
 Torsional shear stress(contd…)
If the building’s center of mass does not coincide with
its center of rigidity, the building will tend to act as if it is
“pinned” at its center of rigidity. It is said to be acted
upon by a torsional moment, Mtorsional, calculated as the
product of the base shear,V, and the eccentricity, e.
This eccentricity is the distance (measured
perpendicular to the base shear’s line of action)
between the centers of mass and rigidity.
Mtorsional = Ve
 Torsional shear stress(contd…)
The rotation about the center of rigidity is resisted by a
torsional shear stress in all members. This torsional
stress is proportional to the distance, r, from the
building’s center of rigidity to the resisting member. In
the equation below, J is the polar moment of inertia of
the resisting members and is most conveniently
calculated as the sum of the moments of inertia taken
with respect to the x- and y-axes.
M torsional
v
J
J  Ix  Iy
 Torsional shear stress(contd…)
The UBC requires that an accidental eccentricity of 5%
(based on the maximum building dimensions at that
level perpendicular to the seismic load) be added to the
actual eccentricity, if any, in design of all buildings, even
those that are symmetrical.
This eccentricity is included to account for accidental
errors in workmanship, uncertainties in the actual
location of the center of mass and rigidity, nonuniform
distribution of dead and live loads, nonuniformities that
result subsequent building modifications, and
eccentricities that develop during an earthquake after
the failure of certain structural elements.
 Negative torsional shear stress
The base shear causes a shear stress that acts in the
same direction in all vertical base members. The
torsional shear stress, however, has different signs on
either side of the center of rigidity. On one side (i.e,
where the resisting element is on the same side of the
center of rigidity as the center of mass) the torsion
increases the stress from the base shear; on the other
side, the stress is decreased. The amount of decrease
is known as negative torsional shear stress. Negative
torsional shear stress should normally be neglected;
that is, it should not be used to decrease the design
capacity of a wall or member.
 Overturning moment
The summation of all moments taken about the base
due to the distributed lateral forces is the overturning
moment, often given the symbol, OTM. If the
overturning moment is large enough, it can reverse the
compression that normally exists in outer columns
caused by the dead and live building loads. Because
footings and concrete walls and columns can be placed
in a state of tension, the overturning moment is more of
a problem for concrete frame and shear wall
construction (which can not tolerate much tension) than
it is for steel frame construction.
 Overturning moment(contd…)
The overturning moment will increase the compressive
stress in outer columns on the opposite side of the
building. Such an increase must be countered by
increasing the thickness of shear walls and using extra
steel reinforcement in concrete columns.

Overturning moments should be calculated for each

building level. The first overturning moment is the sum
of all moments taken about the ground level. This
moment should be used to size footings and to design
the primary outer columns. The overturning moment for
each subsequent floor considers only lateral forces
above that floor. This moment is used to design the
shear walls and other supporting structures at that floor.
 Rigid frame buildings

Before 1965, when the design of structural systems

was still in its infancy, most tall buildings were designed
as rigid frames. In a rigid frame building, columns and
beams were welded together to create a structural grid
that resisted wind and earthquake forces elastically.
Such buildings were expensive to construct because
they used large amounts of material, usually steel, to
keep the stresses in the elastic region.
 High rise buildings

The optimum design for seismic loading often conflicts

with that for wind loading, a significant factor for any tall
building. For an earthquake, the buildings need to be
flexible, even though the full flexibility might be called
upon once in 400 years. However, the full flexibility
might be experienced during large wind storms, say,
once every ten years. The greater flexibility required to
resist large earthquakes makes for unpleasant motions
in wind storms.
 High rise buildings(contd…)

The rigid frame system relies on the bending of

columns and beams for its lateral stiffness. However,
bending is a poor way to tap a structural member’s
strength compared to axial loading.

A tube building resists lateral forces in a radically

different way from a rigid frame building. The tube is
like a giant box beam cantilevering out of the ground.
Axial forces in the columns mainly resist the tendency
to move laterally.
 High rise buildings(contd…)

In order to economically design for increasing

numbers of storeys; different flexible structural systems
were developed. The most general systems are

1. Frames with bracing in the core, which creates a

stiff vertical truss, good for buildings upto 30 to 40
storeys
2. Framed tubes, good for up to 60 or 70 storeys

3. Diagonally braced tubes, good for up to 100 or 120

storeys