Engineering Structures 22 (2000) 201–209

www.elsevier.com/locate/engstruct

Displacement-based seismic design of buildings—theory

*
M.S. Medhekar, D.J.L. Kennedy
Department of Civil and Environmental Engineering, The University of Alberta, Edmonton, Alta., Canada T6G 2G7

Received 27 April 1998; received in revised form 7 August 1998; accepted 7 August 1998

Abstract

The conceptual basis of the spectral acceleration-based design method currently used in seismic codes is reviewed and its limi-
tations are discussed. An alternative method that uses displacements as the basis for the design procedure is then presented. Its
conceptual basis for elastic and inelastic seismic design and its application to single-degree-of-freedom and multi-degree-of-freedom
structures are reviewed. The effects of torsion are considered. The advantages of this method over the spectral acceleration-based
design method are also discussed. To the authors’ knowledge, this and the accompanying paper mark the first application of the
displacement-based procedure to the seismic design of (steel) buildings.  1999 Elsevier Science Ltd. All rights reserved.

Keywords: Buildings; Displacement-based seismic design

1. Introduction normalized design acceleration response spectra, Sa/g,

specified by the NBCC [1] for zones where the zonal
The National Building Code of Canada (NBCC) [1] acceleration and velocity are numerically equal. The
and other design codes [2] use a spectral acceleration- elastic spectral acceleration, (Sa/g)e, corresponds to the
based method for seismic design. Consider a structure estimated period, Te. The elastic base shear that a SDOF
that may be represented by a single-degree-of-freedom system of mass M attracts is
(SDOF) system of mass, m, elastic period, Te, and damp-
ing ratio, ␰. The period, Te, is estimated from empirical Ve ⫽ Mg(Sa/g)e (1)
equations derived from experiments on structures princi-
pally designed for high seismic zones. Fig. 1 shows the If the SDOF system is provided with strength Vi that

Fig. 1. Design acceleration response spectra.

* Corresponding author. Tel.: 001-403-492-1906; Fax: 001-403-

492-0249.

0141-0296/00/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 1 - 0 2 9 6 ( 9 8 ) 0 0 0 9 2 - 3
202 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209

is less than Ve but has sufficient ductility, it responds with the serviceability limit state. At the ultimate
inelastically as shown in Fig. 2 (assuming that the initial limit states, the deformations are likely to contribute
period remains equal to Te). The NBCC [1] accommo- to the instability of the structure and, as well, the
dates this behaviour by reducing the elastic spectrum by damage, perhaps implying that the building is par-
the force modification factor, R, to arrive at the inelastic tially or completely non-functional or even beyond
design spectrum shown in Fig. 1. The inelastic base repair, may also be considered unacceptable. These
shear is thus limit states are governed by deformations. Further-
more, it may be easier to define failure of a structural
Vi ⫽ 冉冊
Ve
R
⫽ Mg(Sa/g)i (2)
element as a limit on deformation rather than as a
limit on the force.
Therefore, it seems rational to examine a seismic
Inelastic seismic design therefore achieves economy design method wherein displacements are considered at
because the system is designed for a reduced base shear the start of the design process with attention focused on
of Vi rather than Ve. However, ductility is required to deformations to provide a structure that meets the
ensure that this inelastic response is obtained and, as requirements for the several limit states. A serviceability
well, it is implicit that the structural and non-structural limit state on deformations could be applied under mod-
damage consistent with the response must be accepted. erate earthquakes that are likely to occur relatively fre-
The spectral acceleration-based design method has the quently in the life of the structure (by imposing drift
following limitations. limits so that non-structural damage is limited or does
1. An estimate of the fundamental period, Te, is required not occur). To prevent collapse in a major earthquake,
to start the design process because the period of the the ductility demand on the structural elements and the
structure to be designed is not known. Seismic codes overall deformation of the structure would have to be
use empirical expressions for the period based on a controlled. It is suggested this can be achieved more
general description of the structural system and its rationally with a displacement-based rather than an
geometry. These estimated periods are intentionally acceleration-based design method.
less than the measured ones to give a conservative Therefore, the objectives of this paper are: (a) to
design [3,4]. present the theoretical basis of the displacement-based
2. The force modification factor, R, is intended to be a design method, and (b) to show how this method is
simple means of arriving at an inelastic design. Seis- applied, as may be used for the serviceability limit state
mic codes specify values for the factor, R, depending and the ultimate limit states, to the design of buildings
upon the material of construction and the type of modelled as multi-degree-of-freedom systems. The
structural system used. However, these values appear method is applied to the design of concentrically braced
to be arbitrary, are difficult to justify, and do not frames (CBFs) in the companion paper [5].
appear to have been established consistently by
experiment or analysis.
3. Displacements are treated in a somewhat cursory 2. Displacement-based design method for a SDOF
manner and are checked at the end of the design pro- system
cess only. There appears to be a lack of concern about
the implied inelastic displacements when values of The central concept of the method, based on Priestley
R greater than 1.0 are used. Both non-structural and [6], is that a structure is designed for a specified target
structural elements may be deemed unsatisfactory if displacement. The method is illustrated by reference to
they deform excessively under earthquakes associated a single-storey, single-bay CBF that is modelled as a
SDOF system as shown in Fig. 3. A set of elastic dis-

Fig. 3. Single-degree-of-freedom model of a concentrically braced

Fig. 2. Inelastic response of a single-degree-of-freedom system. frame.
 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209 203

placement spectra for different levels of equivalent vis- 7. The base shear capacity required, as shown in Fig.
cous damping is required, as shown in Fig. 4. The set 5, is
is generated by integrating numerically the equation of
motion of an elastic SDOF system subjected to earth-
quake ground motions appropriate to the location of the Vb ⫽ Keff⌬max (4)
structure. The design procedure for a SDOF system is
as follows:
where Keff is the secant stiffness when the structure
1. Estimate the yield displacement of the structure, ⌬Y. behaves inelastically, as discussed in the next section.
For the CBF shown in Fig. 3, this displacement is 8. Select structural elements to provide the base shear
the lateral drift that yields the brace and, if column capacity, Vb. Once the members are selected, the
deformation is neglected or is minimal, can be taken initial design of the structure is completed and the
to be a function of the geometry and material proper- structure can now be analyzed. The elastic stiffness
ties (E, FY) of the brace only. and a revised estimate of the yield displacement are
2. Select an appropriate maximum inelastic displace- obtained.
ment, ⌬in, which depends on the deformation capacity 9. Revision is necessary when the members selected in
of the structural elements. For a CBF with nominal step 8 provide significantly different stiffness and
ductility, ⌬in may be taken as 1.0 ⌬Y, i.e. the ductility strength from that required. The displaced shape and
demand is 2.0. A larger value may be selected for effective damping may be revised and steps 5 to 8
appropriately detailed structures. are repeated until a satisfactory solution is obtained.
3. The maximum displacement of the SDOF system,
⌬max, the sum of the yield displacement, ⌬Y, and the
maximum inelastic displacement, ⌬in, is thus related
to the brace ductility demand assumed in step 1.
4. Select an appropriate value of effective structural
damping, ␰eff, which depends on the ductility level
implied in step 2.
5. The effective period, Teff, corresponding to the
maximum displacement, ⌬max, and the effective
damping, ␰eff, is obtained as shown in Fig. 4.
6. The effective stiffness of the SDOF system, Keff, is

4␲2meff
Keff ⫽ (3)
T 2eff

where meff is the mass of the SDOF system. Fig. 5. Elastic and effective stiffnesses.

Fig. 4. Elastic design displacement spectra.

204 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209

3. Basis of the displacement-based design method ally dissipated by an equivalent linear viscous dashpot.
Therefore, in Fig. 4, the effective damping, ␰eff, is the
When a structure behaves inelastically in the design equivalent viscous damping which comprises damping
earthquake, the substitute structure approach is followed that exists in the structure without inelastic behaviour
in order that elastic displacement spectra can be used in (nominal viscous damping) and the hysteretic damping
the displacement-based design method. In this procedure due to inelastic action. The latter, of course, depends on
[7], an inelastic SDOF system is modelled as an equival- the level of inelastic action in the structure. Shibata and
ent linear elastic analogue having the ‘substitute’ proper- Sozen [9] extended this approach to the design of
ties of: (a) effective stiffness, Keff, as shown in Fig. 5, reinforced concrete frames modelled as multi-degree-of-
(b) effective damping, ␰eff, and (c) effective period, Teff. freedom (MDOF) systems.
The maximum displacement of the inelastic SDOF sys- For design, the lengthened or substitute period is that
tem is estimated by using the substitute properties corresponding to the assumed maximum displacement,
together with an elastic displacement spectrum. ⌬max, and effective damping, ␰eff, that is point ‘c’ in Fig.
Consider the response of an inelastic SDOF system 6. The damping, ␰eff, depends on the ductility demand
with an elastic period, Te, and damping ratio, ␰. Let ⌬Y imposed on the structural elements in the earthquake.
be the yield displacement and ⌬max be the maximum tar- Guidelines for selecting ␰eff for reinforced concrete
get displacement. Considering Fig. 6 where displacement structures are given by Gulkan and Sozen [7] and Shib-
spectra for increasing damping ratios ␰ are shown, there ata and Sozen [9]. Little information is available on
are several paths to reach the maximum displacement, effective damping values for steel structures responding
⌬max, starting from the point ‘o’ at Te and ⌬Y on an elas- inelastically. Newmark and Hall [10] recommend a
tic displacement spectrum [8]: (i) with the same period, damping ratio between 5% and 7% for steel structures
Te, select a lower damping ratio to reach point ‘a’, (ii) with welded connections and between 10% and 15% for
with the same damping ratio, select a longer period, T1, steel structures with bolted connections.
to reach point ‘b’, (iii) select both a longer period, T2, Once the substitute period, Teff, is determined for a
and a larger damping ratio to reach point ‘c’, and (iv) desired maximum displacement, ⌬max, the effective stiff-
select a shorter period, T3, and a lower damping ratio to ness of the SDOF system, which is a secant stiffness to
reach point ‘d’. The first and last alternatives are the maximum displacement, ⌬max, as shown in Fig. 5, is
unrealistic and the third is the most realistic because found from Eq. (3). [If the period, Te, corresponding to
structures responding inelastically exhibit both greater the yield displacement, ⌬Y, is used in Eq. (3), the elastic
damping and a longer period due to softening. Gulkan stiffness, Ke, is obtained.] The inelastic base shear for
and Sozen [7] conducted dynamic tests on SDOF con- the SDOF system is given by Eq. (4).
crete frames and deduced guidelines for selecting the An additional issue that needs to be addressed is the
substitute period and damping. The substitute period was decrease in base shear with an increased value of the
determined experimentally by taking the ratio of the assumed maximum displacement, ⌬max. This is illus-
maximum displacement to the maximum absolute accel- trated in Fig. 7(a) for the SDOF system shown in Fig.
eration and using the relation Sa ⫽ ␻2 Sd, where Sa is 7(b), where ⌬1 and ⌬2 are two values of the assumed
the spectral acceleration, Sd is the spectral displacement, maximum displacement, ⌬max, that are both greater than
and ␻ is the circular frequency. The substitute damping the yield displacement, ⌬Y.
was evaluated by assuming that the energy input to the Consider the design of the SDOF system for a
SDOF system during the earthquake excitation was tot- maximum displacement equal to ⌬i. The effective per-
iod, Ti, is determined from an elastic displacement spec-
trum corresponding to the assumed maximum displace-
ment, ⌬i, and effective damping, ␰eff. The effective
stiffness, Keff, is obtained from Eq. (3) and the inelastic
base shear from Eq. (4). Let Ke be the elastic stiffness
obtained from Eq. (3) for period, Te, that corresponds to
the yield displacement, ⌬Y. Assuming the mass, m, to
remain constant, the ratio of the stiffness, Keff to Ke, is
obtained as

(Keff)i
Ke
⫽ 冉冊
Te
Ti
2
(5)

where Ti is the period corresponding to the assumed

maximum displacement, ⌬i. Fig. 8 shows an idealized
Fig. 6. Use of elastic displacement spectra. design displacement spectrum wherein the spectral dis-
 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209 205

Fig. 7. Reduction in base shear with displacement.

4. Displacement-based design method for a MDOF

system

Consider a MDOF system with n degrees of freedom

as shown in Fig. 9. The displacement-based design
method is applied to this system by first transforming it
into an equivalent SDOF system, as outlined by Calvi
and Kingsley [11], and then using the procedure outlined
in Section 2 for a SDOF system. The MDOF system
is forced to behave like a SDOF system by permitting
displacements of only a single pre-determined or
Fig. 8. Linearized design displacement spectrum. assumed shape. The transformation is based on the fol-
lowing assumptions:
placement is assumed to increase linearly with period. 1. the MDOF system responds harmonically in the
This linearization may be considered acceptable for the assumed shape,
range of periods within which the fundamental period of 2. the base shear developed by the MDOF system and
most low-rise to moderate-rise buildings lies. Using the its SDOF equivalent are the same, and
assumed linear relation between period and spectral dis- 3. the work done by the lateral earthquake force on both
placement, Eq. (5) simplifies to systems is the same.

(Keff)i
Ke
⫽ 冉 冊
⌬Y
⌬i
2
(6)
Let the assumed displacement vector of this MDOF
system be represented by

Using Eq. (4), the ratio of inelastic to elastic base {␦} ⫽ {␦(h,t)} ⫽ 具␦1(t),␦2(t),$,␦n(t)典T (8)
shear is
This vector may be expressed as
Vi ⌬Y
⫽ (7) {␦(h,t)} ⫽ {⌽(h)}Z(t) (9)
Ve ⌬i

Thus, for the linearized design displacement spectrum where ⌽(h) is a spatial (shape) function and Z(t) is a
shown in Fig. 8, the base shear decreases in inverse pro- temporal function. Assuming harmonic response, Eq. (9)
portion to the ratio of the assumed maximum displace- is written as
ment, ⌬max, to the yield displacement, ⌬Y, or conversely
the maximum displacement equals the force modifi- {␦(h,t)} ⫽ Zosin(␻t){⌽(h)} (10)
cation factor, R, times the yield displacement when the
spectral displacement varies linearly with the period. This expresses the first assumption that the shape of
206 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209

Fig. 9. Multi-degree and equivalent single-degree-of-freedom system.

the MDOF system does not change with time and only ai ⫽ ciaeff (13a)
the amplitude of motion varies harmonically. Differen-
tiating Eq. (10) twice with respect to time gives the where aeff is the acceleration of the equivalent SDOF
acceleration vector system. Therefore the lateral inertia force at the ‘i’th
mass is given by
{a(h,t)} ⫽ ⫺ Zo␻2sin(␻t){⌽(h)} ⫽ (11)
⫺ ␻2{␦(h,t)} Fi ⫽ miai ⫽ miciaeff (14)

Thus, the acceleration at each DOF, ai, is proportional By assumption 2, the sum of the lateral forces on the
to its assumed displacement, ␦i. MDOF system is equal to the base shear Vb, thus

再冘 冎
Let the properties of the equivalent SDOF system be
冘 冘
n n n
mass, meff, stiffness, Keff, damping, ␰eff. Also let ␦eff and
Vb ⫽ Fi ⫽ miai ⫽ mici aeff (15)
Vb be the effective displacement and base shear of the
i⫽1 i⫽1 i⫽1
equivalent SDOF system, respectively, where these
properties are to be derived from those of the MDOF ⫽ meffaeff
system. The displacement of the MDOF system can be
normalized by dividing by the effective displacement of From Eq. (15), the effective mass of the SDOF system
the equivalent SDOF system to give is defined as

冘
n
1
{c(h,t)} ⫽ {␦(h,t)} (12) meff ⫽ mici (16)
␦eff
i⫽1

or The lateral force, Fi, at each mass from Eq. (14) may
be expressed in terms of the base shear, Vb, by manipul-
␦i
ci ⫽ (12a) ating Eqs. (12a), (14) and (15) to give
␦eff
mi␦i
Fi ⫽ Vb (17)
冘
As accelerations for harmonic motion are proportional n
to displacements mj␦j
j⫽1
ai
ci ⫽ (13)
aeff The effective displacement, ␦eff, is found from
assumption 3 by equating the work done by the lateral
or force on each system
 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209 207

冘 10. Design the lateral load resisting system (LLRS) for

n

Vb␦eff ⫽ Fi␦i (18) these forces and the appropriate gravity load.
i⫽1 Material design standards are used at this stage to
select members and details that have adequate defor-
Solving Eq. (18) for ␦eff and substituting for Fi from mation capacity as implied in the desired displaced
Eq. (17) results in shape in step 1.
11. With appropriate members selected, a design is now
冘
n
available for analysis. As a minimum, a nonlinear
mi␦ 2i static analysis of the structure under the lateral
i⫽1
␦eff ⫽ (19) earthquake force profile is recommended. This
冘
n

mi␦i analysis accounts for approximately the redistri-

i⫽1
bution of internal forces when the structure is loaded
into the inelastic range. The objectives of this analy-
The effective displacement from Eq. (19) is used to sis are to check the displaced shape against the
enter the displacement spectrum, for the appropriate desired one assumed in step 1 and to assess the
level of effective damping, to find the effective period. strength and deformation demands on the structural
With this period and the effective mass from Eq. (16), elements. In following this analysis procedure, one
the effective stiffness is found from Eq. (3) and the base of two cases may obtain. In the first case, the lateral
shear is given by force profile is applied and the load factor is
incremented until the lateral force is attained. Inelas-
Vb ⫽ Keff␦eff (20) tic action may occur at some locations in the struc-
ture. (For direct-acting bracing, the use of a small
The forces at each mass, Fi, are found from Eq. (17) strain-hardening modulus is required so that the
and the design is effected. deformations are not indeterminate). In the second
case, the lateral forces in the assumed profile are
applied until the desired deformation is reached at
5. Design procedure for a MDOF system a specified location, e.g., the roof level. Inelastic
action does take place in the structure. Obviously,
nonlinear dynamic time history analyses give a more
The design procedure for a generalised MDOF sys-
realistic assessment of the drift and the ductility
tem is:
demands. When the hysteretic behaviour of compo-
1. Select an initial desired displaced shape for the nents is modelled directly, the effective damping
structure, ␦i. For a multi-storey CBF, the desired ratio to be used is that pertaining only to other
drift at each floor level is specified. The brace duc- sources of damping (nominal viscous damping).
tility demand can be calculated from the assumed 12. If the displaced shape and the associated defor-
storey drift based on the approximation that it mation demands on the structural elements from the
depends only on the geometry and material proper- nonlinear analyses are significantly different from
ties of the frame. (This neglects the axial defor- those assumed, the stiffness and strength distribution
mations of the column, which may be appropriate should be revised to obtain the desired result. This
for low-rise structures). Methods for estimating the is most likely to occur when higher modes of
contributions of column deformations to drift are vibration have a significant effect.
discussed in the companion paper [5].
2. Select the effective damping, ␰eff, for the structure.
This depends on the ductility implied with the 6. Torsional effects
desired displaced shape, ␦i.
3. Determine the effective displacement, ␦eff, from Buildings should be designed and constructed, insofar
Eq. (19). as possible, by modifying the frame stiffnesses to mini-
4. Obtain the normalized profile, ci, from Eq. (12). mise the structural eccentricity, defined as the distance
5. Obtain the effective mass, meff, from Eq. (16). measured perpendicular to the direction of seismic load-
6. Determine the effective period of the SDOF system, ing between the centre of mass and the centre of rigidity,
Teff, from an elastic design displacement spectrum at the level being considered. Whatever structural eccen-
for displacement, ␦eff, and damping, ␰eff. tricity remains results in a torsional moment and attend-
7. Obtain the effective stiffness, Keff, from Eq. (3). ant shear forces that are distributed on the basis of equi-
8. Obtain the base shear, Vb, from Eq. (20). librium and frame interstorey drifts to the frames. The
9. Obtain the lateral forces on the MDOF system from adequacy of the frames is checked and they are modified
Eq. (17). as required.
208 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209

Paulay [12,13] describes a design strategy for earth- 3. Empirical equations for estimating the fundamental
quake-induced torsion in buildings which is based on period of the structure for preliminary design of the
the deformation capacity of the structural system and its LLRS are not required.
critical elements rather than on torsional strength, as is 4. The selection of a displaced shape at the start of the
consistent with the displacement-based design approach. design process forces the engineer to consider the
When inelastic action occurs, based on what is essen- configuration of the LLRS and the drift tolerance for
tially a lower bound approach for elements displaying the non-structural elements. The displaced shape may
elasto-plastic behaviour, Paulay [12] assumes that these be linked explicitly to the member ductility demand,
elements parallel to the lateral force have yielded due to as is the case in the CBF. The drift of non-structural
translation and therefore that the torsional moment can elements at various levels of damage can be obtained
only be resisted by the perpendicular elements whose from experiments and used directly in design.
resistances must form a couple. If a couple cannot be 5. The empirical and somewhat arbitrary force modifi-
formed, the system is considered to be torsionally unre- cation factor, R, used in the spectral acceleration-
strained. For torsionally restrained systems, the elements based design method, is not needed.
resisting torsion are likely to behave elastically. In any
case, together with the assumption of rigid body rotation
of the floors, the other assumptions reduce the system 8. Closure
to a statically determinate one. All the frames in the
direction of the earthquake are at their yield capacities. The displacement-based design method, as outlined in
The torsional couple formed by the earthquake force act- Sections 2 and 4, has been applied to the seismic design
ing at an eccentricity measured perpendicular to the of RC bridge columns [14] and MDOF bridge structures
earthquake force from the centre of mass to the centre [11], respectively. To the authors’ knowledge, this
of strength is resisted by the perpendicular elements. The method has not been applied to the design of steel build-
torsional deformations of these elements are taken to be ings yet. The concept of replacing a MDOF building
proportional to the distance from the centre of strength where the masses and stiffnesses are in series by an equi-
(the location where the resultant of the resistances of valent SDOF system, presented by Calvi and Kingsley
the inelastic translatory elements acts). Because of the [11] for bridge structures where they act in parallel, also
twisting displacements, the centre of mass has an introduces the following issues.
additional displacement above the translational one and 1. Selection of an appropriate displaced shape and its
some structural or non-structural element becomes the effect on the design base shear, the static lateral force
critical one depending on the drift criteria established. profile, and the member design requirements.
Both effects must be considered. Because of the elastic 2. The effect of axial deformation of the columns on the
behaviour of the elements resisting the torsion, it is displaced shape.
likely that the increase in the displacement of the centre 3. The greater cumulative effect of the gravity loads on
of mass due to torsion is small. Therefore, a logical the building (P ⫺ ⌬ effect).
approach is to base the initial design on the translation 4. The necessity of controlling the interstorey drift
of the centre of mass and to modify this assumption sub- (storey ductility demand), particularly in the upper
sequently as required. A simple design strategy for tor- storeys, due to the influence of higher modes of
sion therefore exists. vibration.
In the accompanying paper [5], this method is applied
7. Advantages of the displacement-based design to the design of a two-storey and an eight-storey building
method with concentrically braced steel frames as the lateral load
resisting system.
The displacement-based design method is considered
to offer the following advantages over the spectral accel-
eration-based design method. Acknowledgements
1. Displacements play a major role at the preliminary The financial support of the Natural Sciences and
design stage itself resulting in good control on dis- Engineering Research Council of Canada for the work
placements over the entire design process. Target dis- on both papers in this set is gratefully acknowledged.
placement criteria are selected for the serviceability
and ultimate limit states and thus damage control is
achieved directly. References
2. The strength and stiffness of the lateral load resisting [1] Canadian Commission on Building and Fire Codes. National
system (LLRS) are chosen to satisfy the desired building code of Canada. Ottawa, Ontario: National Research
deformation criteria. Council of Canada, 1995.
 M.S. Medhekar, D.J.L. Kennedy / Engineering Structures 22 (2000) 201–209 209

[2] International Conference of Building Officials. Uniform building look at the substitute structure approach. Earthquake Spectra
code, vol. 2. Structural engineering design provisions. Calif.: 1994;10(2):319–31.
Whittier, 1997. [9] Shibata A, Sozen MA. Substitute-structure method for seismic
[3] Building Seismic Safety Council. National earthquake hazards design in R/C. Journal of the Structural Division, ASCE
reduction program recommended provisions for the development 1976;102(ST1):1–18.
of seismic regulations for new buildings. Washington, DC.: [10] Newmark NM, Hall WJ. Earthquake spectra and design. Oakland,
Building Seismic Safety Council, 1994. Calif.: Earthquake Engineering Research Institute, 1982.
[4] Applied Technology Council. Guidelines for the seismic rehabili- [11] Calvi GM, Kingsley GR. Displacement-based seismic design of
tation of buildings (75% complete draft), Report ATC-33.03. multi-degree-of-freedom bridge structures. Earthquake Engineer-
Redwood City, Calif.: Applied Technology Council, 1995. ing and Structural Dynamics 1995;24:1247–66.
[5] Medhekar MS, Kennedy, DJL. Displacement-based seismic [12] Paulay T. Displacement-based design approach to earthquake-
design of buildings—Application. Engineering Structures (this induced torsion in ductile buildings. Engineering Structures
issue). 1997;19(9):699–707.
[6] Priestley MJN. Myths and fallacies in earthquake engineering— [13] Paulay T. Seismic design for torsional response of ductile build-
conflicts between design and reality. Bulletin of the New Zealand
ings. Bulletin of the New Zealand National Society for Earth-
National Society for Earthquake Engineering 1993;26(3):329–41.
quake Engineering 1996;29(3):178–98.
[7] Gulkan P, Sozen MA. Inelastic response of reinforced concrete
[14] Kowalsky MJ, Priestley MJN, MacRae GA. Displacement-based
structures to earthquake motions. Journal of the American Con-
design of RC bridge columns in seismic regions. Earthquake
crete Institute, 1974, paper no. 71-41, December, 604–10.
Engineering and Structural Dynamics 1995;24:1623–43.
[8] Bonacci JF. Design forces for drift and damage control: A second