RECENT TRENDS IN THE SEISMIC

DESIGN OF REINFORCED CONCRETE

BUILDINGS

Prof. Tamás Paulay

To enable acceptable seismic displacement demands imposed by earthquakes to be estimated, the

displacement capacity of a structure needs to be known. This is controlled by selected performance
criteria. It is postulated that the assignment of fractions of the required seismic strength to
components of the system may be arbitrary. This enables the displacement capacity of a system to
be evaluated without the knowledge of its strength. A redefinition of some traditionally used
structural properties, such as stiffness, is a prerequisite of applications. The study of a reinforced
concrete mixed framewall system illustrates rationale, acceptable extents of approximations and
extreme simplicity of application.

Keywords: strain, curvature, displacement, ductility, detailing, strength, stiffness

1. INTRODUCTION

Earthquake shaking imposed on buildings represents perhaps the greatest challenge that a structural
designer may need to face. In many countries seismic structural design criteria override all others
that relate to reinforced concrete building systems. Limit state design procedures which consider
structures subjected to gravity loads, wind forces and environmental effects, address hypothetical
failure states. However, in regions of significant seismicity the attainment of displacements close to
those of a failure state, is almost a certainty. In this scenario safety relies on a small probability of
the occurrence of a major earthquake that could create conditions corresponding to an ultimate limit
state during the probable time span for the use of the structure.
In contrast to requirements to be satisfied in regions of no or small seismicity, seismic structural
design needs generally to address inelastic dynamic response, significant repeated displacement
reversals, inevitable stiffness and possible strength degradations. The detailing of the reinforcement
in potential plastic regions of the structure is, therefore, of overriding importance. This review
attempts to highlight some findings derived from relevant theoretical and experimental research
conducted particularly in New Zealand over the last three decades.
Traditional techniques, adopted in early seismic code provisions, attempted to provide adequate
strength, in terms of lateral force resistance, largely based on the dynamic response of elastic
systems. It was recognized that for rare seismic events lateral design forces can be significantly
reduced if proper allowance is made for hysteretic damping. Some associated damage is, however,
inevitable. This presented a major challenge to researchers to identify the sources of large
deformation capacities in a composite structure comprising essentially brittle concrete and
adequately ductile steel. An astute choice of a kinematically admissible plastic mechanism needs to
be made. To ensure that only the suitably chosen mechanism can be mobilized during a large
earthquake, a hierarchy in the relative strengths of components is also required. These concepts are
embodied in the philosophy of "capacity design" (Park and Paulay, 1975, Paulay and Priestley, 1992).
More recently emphasis in seismic design strategies was placed on satisfying identifiable performance
criteria. Instead
of providing only seismic strength, the importance of realistic predictions of displacement capacities
of structural systems, associated with specific performance criteria, is being recognized.
To aid simplifications, without which the adoption of new concepts in seismic design cannot be
expected, deliberate approximations need to be made. These are considered to be compatible with
the inevitable crudeness of the prediction of magnitudes of earthquake-induced displacement
demands.
Certain redefinitions of structural properties, for example the relationship between strength and
stiffness, may be viewed by some readers as being controversial.

2. A DISPLACEMENT FOCUSEDSEISMIC DESIGN APPROACH

2.1 Terminology used

In this study of earthquake-induced displacements of buildings, reference are made to the structural
system.
A structural system comprises lateral force-resisting elements, generally arranged in orthogonal
directions. Typical elements are bents of ductile frames or interconnected walls in the same plane.
Due to torsional effects, elements of the system may be subjected to different displacements.
A lateral force-resisting element may comprise several components. The components will, however,
be subjected to identical lateral displacements. Typical components are beams, columns, beam-
column subassemblages or individual walls.

Fig. 1 Nomenclature

Fig. 1 illustrates typical elements and components of an example structural system. In the technical
literature the terms component and element are often used to describe the same part of the system.
Distinct differentiation between components and elements are made in this study.

2.2 Displacement limitations

Following currently accepted seismic design aims, maximum displacements imposed on reinforced
concrete elastoplastic systems are deemed to be controlled by:
- The displacement capacity of critical components of the system corresponding with the adopted,
i.e., codified, quality of detailing for construction.
- The magnitude of the storey drift, (the lateral displacement of a level relative to that of an adjacent
level) satisfying the specific performance criterion, specially chosen for a building system.
- The more severe limit may then establish the target displacement capacity of the ductile system.

2.3 The tools of displacement estimates

Traditionally the structural design process starts with experience-based estimates of dimensions,
particularly component sizes which are likely to satisfy functional requirements of the building. Once
this information, based on architectural and engineering perspectives, is available, with the
knowledge of the material properties, strength-independent displacement estimates, adequate for
purposes of seismic design, can be readily made.
Fig. 2 Flexural strength-curvature relationships for a wall section

2.3.1 Nominal yield curvature

A fundamental property of a structural component is the nominal yield curvature at its critical section
or sections. This will define its response in the elastic and postelastic domain of behaviour when it is
subjected to monotonically increasing displacements. In reinforced concrete members, it is more
realistic to base curvature estimates on quantifiable section properties, rather than on assumed or
recommended (Paulay and Priestley, 1992) values of flexural rigidity, EcIe, where Ec is the modulus of
elasticity of the concrete and Ie is the second moment of effective sectional area of the cracked
component. In seismic design it may be assumed that extensive cracking will occur over the full
length of components.
Fig. 2 shows flexural strengthcurvature relationships for a typical structural wall section. With some
experience the neutral axis depth associated with the development of steel yield strain at the
extreme tension fiber can be readily estimated. As stated, for purposes of seismic design, a high
degree of precision in the estimation of component properties is not warranted. However, if
necessary, this estimate can be subsequently verified once details of the flexural reinforcement
provided are known. For a given steel tensile yield strain, åy, and the location of the neutral axis
depth, xDw, the curvature at the onset of yielding is established as jy = ey /(xDw), where Dw is the
overall depth of the section. When details of the reinforcement are known, the associated yield
moment, My, may also be evaluated. The need for this will, however, seldom arise. The designer will
primarily address the nominal (ultimate) flexural strength, Mn, at a section, as constructed, when
strength requirements for the components are known.
For seismic design purposes the bilinear simulation on the nonlinear moment-curvature relationship,
as shown in Fig. 2.b, is convenient and adequate. With linear extrapolation this leads to the definition
of the nominal yield curvature
jy= (Mn /My ) j`y= hey/Dw (1)
where the coefficient h = (Mn/My )/x recognizes the ratio of the nominal to yield strength in flexure
and the relative position of the neutral axis. Effects of strain hardening with increasing curvature
ductility may also be included. These, however, are not shown here.
Extensive studies of a variety of sections, conducted at the University of Canterbury, confirmed
previous findings (Priestley, Kowalsky, 1998) that for specific types of members, such as walls or
beams, the variation of the value of the parameter h is relatively small. For example the amount of
reinforcement used at a section hardly affects the nominal yield curvature. Neither do moderate axial
compression loads, commonly encountered in structural walls, affect nominal yield curvature to any
significance. However, the effect on flexural resistance is very pronounced. Fig. 2.b, where the
(cylinder) compression strength of the concrete is denoted as f `c , illustrates this feature.
A message of this review of the nominal yield curvature is, that it should be considered as a material
property. It is insensitive to, and for design purposes essentially independent of, section strength.
The definition presented here contradicts with the widely used terminology, whereby jy=My /(EcIe ).

2.3.2 Nominal yield displacement

An approach, similar to that used in defining the nominal yield curvature at a cracked reinforced
concrete section, based on bilinear simulation of force-displacement relationship for a component,
may be utilized. Fig. 3 shows familiar features of this simulation. The strongly nonlinear response of
component i from the onset of cracking till the development of its nominal strength, Vin, is of little
interest with respect to estimations of displacements in the inelastic range of behaviour. The shaded
area in Fig. 3 indicates essentially linear response after the occurrence of repeated displacements not
exceeding that associated with the yield strength, Vy, and the yield curvature, j`y of the component.
Hence the nominal yield displacement, Dy, an essential reference value, may be based on the
nominal yield curvature, j , at the critical section. It is important to note that the nominal yield
displacement, Dy, is also strength-independent!
In this study, only conservatively estimated flexural deformations were considered. If refinement
appear to be necessary, other sources of distortions, such as due shear and anchorages, may be
readily included. Subsequent examples will illustrate applications.

Fig. 3 Bi-linear simulation of force-displacement relationship

2.3.3 Component stiffness

An important conclusion, drawn from the bilinear simulation of the force-displacement relationship
for a component, is the definition of its stiffness, stated in Fig. 3, Contrary to usage in traditional
design practice, the stiffness of a reinforced concrete component, with given dimensions and material
properties (ey), is proportional to the strength which the designer will eventually assign to it (Paulay,
2001a).
Fig. 3 clearly shows the range of strengths, i.e., when Vy<V<Vn, over which stiffness so defined
would significantly underestimate displacements. In terms of ductile structural response this
transitional range of behaviour is in general of little interest.
The bilinear simulation shown in Fig. 3 allows the displacement ductility, applicable to component i to
be more realistically quantified as
miD = Dm/Diy (2)
The displacement ductility, mD, of a system, to be defined subsequently, is an essential parameter of
current strength and displacement-based seismic design procedures.

2.4 Freedom in the assignment of strengths

As stated previously, the nominal yield curvature at the critical section of a component with given
dimensions, jiy, and hence its nominal yield displacement, Diy, may be considered in seismic design to
be independent of the nominal strength. Therefore, fractions of the total required nominal strength of
a system may be assigned arbitrarily to its components or elements (Paulay, 2000). Implications of
this freedom in assigning component or element strengths, are thus:
- Irrespective of the strengths assigned to components, characterized by biliner force-displacement
relation ships, they will commence yielding when the imposed system displacement approaches the
relevant nominal yield displacement.
- Simultaneous onset of yielding of components of elements, such as walls, with different overall
depths, and hence different aspect ratios, is not possible.
- Lateral force-resisting components of a system with different nominal yield displacements, when
subjected to identical translational limit displacements, will be subjected to different displacement
ductility demands.
- A structural system, comprising elements with different nominal yield displacements, does not
exhibit a distinct yield displacement. Existing definitions of the yield displacement of systems are
often ambiguous. Hence a redefinition is required.
- To ensure that all components of a ductile system will perform satisfactorily, the displacement
capacity of the system should be restricted to that of its component with the smallest displacement
capacity.
Fig. 4 The model of a wall system

Arbitrariness in strength assignment imparts to the astute designer the ability to chose from a
number of possible, viable and appealing solutions. To illustrate the relevance of the structural
features listed above, a somewhat idealized system, deliberately made simple, is studied here briefly.
Fig. 4.a shows the plan of a structure in which lateral forces in the Y direction are resisted by five
rectangular reinforced concrete cantilever walls of identical heights but different length, Di. Slender
components with very small lateral strength, intended for the transmission of gravity loads only, are
not shown in Fig. 4.a. A single large wall element is provided to resist seismic forces in the X
direction.
Design approaches, based on traditional definitions of component stiffness, would assign nominal
wall strengths, i.e., base shear, in proportion to D. Under identical element translations this would
require the reinforcement ratios in Walls (1) and (3) to be 43% larger than for Walls (4) and (5).
However, if nominal strengths are made proportional to say, D,all walls would require the same ratio
of vertical reinforcement. This would enable identical arrangement of bars to be used in all walls.
Wall strengths so provided would. however, result in this case in a strength eccentricity of 0.095a.
With some 6% reduction of the strength of Wall (1) and a corresponding increase of Walls (4) and
(5), even this relatively small eccentricity would be eliminated. Thereby torsional phenomena,
affecting element seismic displacement demands primarily in the postyield domain of response,
would be negligible (Paulay, 2001b), and would thus not need to be addressed. It may be shown that
the strength-dependent stiffness eccentricity, would also be rather small (0.21a, i.e., approximately
4% of the length of the plan).
The bilinear simulation of the translational behaviour of the system and its 5 components is
presented in Fig. 4.b. Displacements plotted are normalized in terms of the nominal yield
displacement of Walls (1) and (3), taken as unity. The force-displacement modelling of the elements
with different strength and stiffness properties, demonstrates that:
- During monotonic displacements of the system the sequence of component yielding, from that of
Wall (1) to Wall (5), is independent of wall strength. Nominal yield displacements of the walls are
inversely proportional to wall lengths, Di.
- The superposition of the bilinear response of elements results in a nonlinear response of the
system.
- Assuming bilinear simulation also for the system, would be justified. This assumption allows the
equivalent nominal yield displacement of the system to be expressed as
Dy=SVin/S(Vin/Diy )=SVin/Ski (3)
In this example Dy=1.00/0.879=1.14 displacement units. It is to be used as a reference value to
enable the displacement ductility capacity of the system to be gauged. It is seen in Fig.4(b) that at
this displacement Walls (1) and (3) would have entered the inelastic domain, while the other walls
are still elastic.
- As stated previously, the displacement capacity of the system should be limited to that of it critical
elements, i.e., Walls (1) and (3). In this example it is assumed that appropriate detailing of the
reinforcement in all walls, allows a wall displacement ductility capacity of miD =4 to be relied on.
Therefore, the displacement capacity of the system is to be limited to that of Wall (1), Du=4x1=4
units. The displacement ductility capacity of the system, often referred to as global ductility, is thus
to be restricted to mD =4.0/1.14=3.5. Contrary to current codified methods, the global ductility of a
system should not be assumed, but made dependent on the displacement capacity of its critical
element.
This simple example is presented in support and illustration of the claim (Paulay, 2000) that the
displacement capacity of systems can be established before its required strength is determined.

3. AN EXAMPLE FRAME_WALLSYSTEM

An attractive mode of seismic resistance in medium to high rise buildings may be achieved with the
use of interacting cantilever walls and frames, extending over the full height. Before the arrival of the
computers, these systems represented one of the most formidable challenges to the analytical skills
of structural engineers. With the appreciation of some limitations of the uses of bilinear modelling,
generally insignificant in a nonlinear seismic scenario, both required strength and displacement
capacity can be readily estimated.
This example intends to demonstrate how, already at the preliminary stage of the design, such
simple behaviour-based predictions can be made, even for such a complex structure. The strategy
used exemplifies a deterministic design philosophy, whereby the designer simply `tells the structure'
what it should do in the event of a major earthquake.

Fig. 5 A ductile frame-wall system

Displacement compatibility under lateral force is assumed to be assured by infinitely rigid floor
diaphragms. The traditional design approach to the distribution of seismic strength to different
elements, based on elastic behaviour, was extensively studied (Paulay, Priestley, 1992). However,
issues of displacement predictions, relevant to this type of ductile structures, are less familiar.
A prototype structure, shown in Fig. 5.b, will illustrate several postulated, yet unconventional, design
concepts. A symmetrical 12 storey reinforced concrete building comprises seven identical frames and
two cantilever walls, each with an aspect ratio of Awr=7.1. Dimensions are expressed in terms of the
total height, h, of the structure. For modelling purposes the 9 elements are condensed into two
elements, one com prising 7 frames and the other 2 walls, respectively, as seen in Fig. 5.b. A typical
seismic design force pattern is shown in Fig. 5.a. These equivalent static forces lead to moment and
shear force patterns presented in Figs. 5.c and d, respectively. Possibilities for the assignment of the
lateral forces to the two very different elements, and associated displacement limits, are of prime
interest. It is restated that, within rational limits, fractions of the overturning moments, M, and
associated storey shear forces, Vs, to be sustained by the chosen ductile mechanism of the system,
may be assigned to the wall and frame elements in an arbitrary manner.
The mechanism chosen for this structure comprises plastic hinges at the base of the walls and the
columns and at each ends of the beams. Other regions of the components will be provided with
sufficient reserve strength to ensure that no inelastic deformations of significance would occur while
ultimate target displacements are being developed.
To illustrate the principal steps of displacement estimates, familiar simple expressions, reflecting
component behaviour, will be used. Benchmark deformations are independent of strength provided.
Displacements will be able to be compared with component dimensions, shown in Fig. 5.b. For design
practitioners the latter are more meaningful quantities. The symbols Db, Dc and Dw define the overall
depths of beams, columns and walls, respectively.
3.1 The assignment of relative element strengths

A consideration, encountererd in practice, could be a limitation on the transmission of relatively large

overturning moments by the walls. These moments may impose excessive demands on the
foundations. Significant deformations of the foundation structure could adversely affect displacement
limits applicable to the superstructure.
One attractive solution could result in beams, which, at the ultimate limit state, would be subjected
to identical seismic strength demands. This implies that beams at all levels, carrying also identical
gravity loads, could have the same dimensions and identically detailed reinforcement. Thereby the
storey moment capacities, i.e., the product of the lateral displacement-related nominal storey shear
forces of the frame element and the storey height, would be made identical. The corresponding shear
forces across the upper 9 storeys of the frames, associated with the identical nominal strengths of
the beams, would be the same. It is transparent that the major motivation in this search for suitable
distribution of component strengths is practicality in construction.
The development of identical nominal storey shear strengths over the full height of the frame
element, shown in Fig. 5.d, corresponds to the application of a single lateral force at roof level. In
this example it was judged that this single roof level force, shared by 7 frames, could be 35% of the
total base shear, i.e., 5% per frame. As Fig. 5.d shows, with this decision, the contributions of the
walls and frames to the resistance of storey shear forces, Vs, are determined. Subsequently
diaphragm forces, associated with this allocation of strengths, with appropriate magnifications for
dynamic effects, to be examined subsequently, will need to be carefully considered.
The corresponding contribution of the frame element to the resistance of overturning moments, M,
will thus increase linearly from level 13 to its maximum, 0.35hVb, at the base. The remainder of the
overturning moments, to be resisted by the walls, is shown by the shaded area in Fig. 5.c. With this
choice the total overturning moment at the base of the system would be resisted by the two
condensed elements in close to equal proportions.

3.2 Displacement considerations

3.2.1 Wall deformations

It was established that yield deformations are proportional to the yield strain of the reinforcing steel
used. To simplify subsequent expressions and to provide a better feel for relative magnitudes, it is
assumed that in this example ey=0.002, i.e., the yield strength of the steel is 400 MPa. As Fig.5(c)
shows, with the approximate location of zero wall moment, the effective height of an equivalent
cantilever wall may be defined by he=0.63h. For seismic design purposes the variation of wall
moments over this height is assumed to be linear. The nominal yield curvature of the wall base, as
stated in Eq. (1), is thus
jwy=1.8×0.002/(0.14h)=0.0257/h (4a)
where for typical rectangular wall section h =1.8. This leads to the nominal yield displacement of the
wall at the effective height to be
Dwy=jwyh/3=0.0257×0.632h/3=3.4×10 _3h (4b)
The slope of the wall at the same height, he, is
qwy=jwyhe/2=0.0257×0.63/2=8.1x10_3 rad. (4c)
These are important benchmark values because the deformed shape of the entire structure will be
controlled by that of the wall element.

3.2.2 Storey deformations of frames

Nominal interstorey yield deformations of frames in which the yielding of columns is suppressed by
appropriate capacity design procedures, referred to in the introduction, originate primarily from the
nominal yield curvatures in the potential plastic hinges of beams. Additional elastic deformations will
occur due to shear effects in beams, columns and joints, and flexural rotations of columns. Because,
compared with nominal yield rotations of beams, the contribution of elastic deformations, listed
above, are relatively small, these may be estimated. Details are not given here. It has been shown
(Priestley, 1998) that a reasonable estimate, particularly for seismic assessments, of the nominal
yield drifts of storeys in frames, is
qfy=0.5 eyAb r= 6.15ey=0.0123 rad. (5)
where Abr is the aspect (span/depth) ratio of the beams, in this example taken as 12.3.
3.2.3 Relationships between ductilities of elements

A comparison of the wall rotation and the nominal yield drift of the frame at level he, obtained from
Eqs. (4c) and (5), shows that, at the nominal yielding of the wall base, the drift in the critical storey,
will be only qwy/ qfy = 8.1/12.3 = 66% of the nominal yield drift of the frames.
The generally accepted maximum seismic storey drift, associated with the ultimate limit state, is in
the order of 2.5%, i.e., 1 in 40. With this limit the associated displace-ment ductilities imposed in the
walls and the frames may now be compared with their displacement capacities. Using Eq. (4c) the
acceptable plastic rotation at the wall base should be limited to
qwp = 0.025 - 0.0081 = 0.0169 rad. (6)
The corresponding plastic wall displacement at level he will be thus
Dwp=(he_0.5 lp) qwp =0.97he qwp = 0.0103h (7)
where the length of the equivalent plastic hinge was found (Paulay and Priestley, 1992) to be lp =
0.0324h.
Therefore, the displacement ductility demand on the wall should be limited to
mwÄ = (Dwy+Dwp) /Dwy = 0.0137/0.0034 » 4 (8)
This value can be readily achieved with appropriate detailing of the plastic hinge region at the wall
base.
The corresponding storey displacement ductility imposed on the frame element in the vicinity of the
effective height is
qwmax/ qfy = 0.025/0.0123 » 2 (9)
This limited ductility capacity can be very easily achieved.
The deformed shapes of the structure and its elements, associated with these benchmark
displacements, are summarized in Fig. 5.e.

3.2.4 Relationships between ductilities of elements and that of the system

This preliminary study of an example frame-wall structural system enables the ductility relationships
between elements, i.e., walls and frames, and the total system to be clarified. These relationships
are not likely to change substantially as the detailed design progresses. Fig. 6 illustrates the bilinear
modelling of force-displacement relationships for the elements and the system.
The relationship between displacement ductilities, with reference to the effective height, he, is very
similar to that shown for a wall system in Section 2.4 and Fig. 4. Hence the specific values,
applicable to the example framewall system, can be presented here.
The stiffness of the elements, based on the definition given in Fig. 3, and the total unit base shear
force, are:
kwall = Vwn/ Dwy = 0.65/0.00340 = 191
kframe = Vfn/ Dfy = 0.35/0.00775 = 45
Sk i = the system stiffness = 236
Therefore, from Eq. (3), the equivalent nominal yield dis-placement of the system, Dy, a rather
important reference value, is
Dy = SVin/Ski = 1.00/236 = 4.24h x 10_3 (10)
i.e., approximately 0.67% of the effective height. The corresponding displacement ductilities are,
therefore:
Walls mwD = as given by Eq. (8) » 4.0
Frames mfD = 0.0137/(0.63 × 0.0123) » 1.8
System mD = 0.0137/0.00424 » 3.2
It is seen thus that the seismic strength, Vb, of the system, corresponding with the specified
acceleration spectrum, should be based on a system (global) displacement ductility capacity of 3.2.
This procedure should protect the critical element (the walls) against demands exceeding a
displacement ductility of 4. As Fig. 3 implies, it will be appreciated that over the approximate range
of displacements, 0.5<D/Dy<2.5, multilinear simulation represents tangents to the continuous
nonlinear forcedisplacement response of a system, often used in pushover analysis techniques.
Fig. 6 Bi-linear modelling of the force-displacement relationship in a wall-frame system

3.2.5 Special features of frame-wall systems

When only cantilever walls provide lateral force resistance, as in Fig. 4, their displacement ductility
capacity need to be significantly curtailed when aspect ratios, Awr=h/Dw, exceed approximately 5.
Wall slopes near roof level, associated for example with a system displacement ductility capacity of
3.2, may become excessive.
In ductile frames, exceeding approximately 3 storeys, displacement capacities are generally
governed by limits on interstorey drift, rather then component deformation capacities. Moreover,
during the elastoplastic dynamic response, frame deformations may become sensitive the effects of
higher modes of vibrations.
To a large extent these shortcomings of deformation behaviour of both walls and frames, may be
eliminated when such elements are coupled within the building by rigid floor diaphragms (Paulay,
2002).
As stated, system deformations are controlled by those of the walls. The nominal flexural strength of
the walls provided at floor levels above the base should be, in accordance with the principles of
capacity design, significantly in excess of that indicated in Fig. 5.c. This procedure should ensure that
wall sections above the potential plastic region at the base will be subjected, at worse, to very small
curvature ductility demands as they are expected to respond in the elastic domain.
When walls, more slender than those shown in Fig. 5.b, are used, for the sake of drift control the
approximate location of zero wall moment must be lowered. This can be achieved by assigning a
greater share of the base shear, Vb, to the ductile frames. For example an equal sharing by the wall
and frame elements of the base shear resistance, would enable the effective height, he, to be
reduced to approximately 0.38h. Thiswould enable a rather slender wall, with an aspect ratio of Awr »
12, to be used without exceeding a drift limit of 2.5%.
This issue indicates also how the designer may assign strength to elements to control critical storey
drift.
It is emphasised that the restriction of plastic hinge formation only to the base of a wall element will
ensure that storey displacement, i.e., drift, imposed on frames will be very similar over the height, h.
However, dynamic effects on the elastic portion of the walls should be expected to increase
significantly local flexural and shear strength demands recorded in Figs. 5.c and d. To safeguard
walls against unexpected ductility demands above the base region, design action, based on lateral
static design forces, need to be increased by appropriate dynamic magnification factors (Paulay and
Priestley, 1992), details of which are beyond the scope of this presentation.

4. CONCLUDING REMARKS

The presented review attempted to highlight some findings on the seismic design of reinforced
concrete buildings derived from relevant theoretical and experimental research conducted particularly
in the New Zealand over the last three decades.
- To satisfy the intents of performance-based seismic structural design, the importance of more
realistic predictions of target displacement capacities should be recognised. For reinforced concrete
structures, addressed here, such displacement limits can be readily and realistically predicted in a
rather simple way without the knowledge of the eventual seismic strength required. Therefore,
displacement estimates made during the preliminary stages of the design, can immediately expose
undesirable features of the contemplated structural system.
- The use of a number of simple principles, often overlooked or ignored in seismic design, was
demonstrated. These include: (a) The stiffness of a reinforced concrete component may be
considered to depend also on the nominal strength eventually assigned to it. Therefore, element or
system stiffness cannot be a priori assumed. (b) The nominal yield curvature of a reinforced concrete
section, which represents a characteristic strain pattern, and displacement of a component
associated with it, are insensitive to the flexural strength of the section. (c) Because deformation
limits applicable to components of a ductile system, exposed to typical seismic moment patterns, are
insensitive to component strength, the latter can be arbitrarily assigned to them. This enables the
astute designer to distribute the required total seismic strength among components so that more
economical and practical solutions, satisfying also stipulated displacement limits, are obtained.
- The estimation of displacement capacities of components of a system enables the critical
components to be identified. Hence, instead of assuming global ductility factors for structural
systems, their displacement and hence ductility capacity should be made dependent on that of the
critical component. Such relationships can be established before strengths are assigned to
components.
- The approach, illustrated with the aid of an example wall and a frame-wall structure, can be readily
incorporated into existing strength-based seismic design methods. Its major appeal relates, however,
to displacement-based seismic design strategies.
- In the seismic design of ductile structures, bilinear modeling of force-displacement relationships for
both components and the system may be considered adequate.
- Moment patterns used for the estimation of deformations of elastic elements are based on typical
static lateral force patterns. The fact that force patterns encountered during seismic response may be
very different, is not considered to invalidate the approach used for the estimates of displacements in
systems dominated by the behaviour of reinforced concrete walls.
- No attempt in this presentation was made to estimate displacement demands. It is the designer=s
responsibility to establish, with the use of a force-based or a displacement-based strategy, the level
of seismic strength that will ensure that, for a given seismic scenario, the displacement capacity of
the system is not likely to be exceeded.
- The approach presented is design rather than analysis oriented. It is based on very simple
principles. It is a useful tool in the hands of a designer, which enables, even for mixed structures
considered in this study, efficient, practical and simple solutions to be obtained.

5. REFERENCES

Park, R., Paulay, T., (1975), "Reinforced Concrete Structures", John Wiley and Sons, New York, 769
p.
Paulay, T., Priestley, M.J.N., (1992), "Seismic Design of Reinforced Concrete and Masonry Buildings",
John Wiley and Sons, New York, 767p.
Paulay, T., (2000), "A simple displacement compatibility-based design strategy for reinforced
concrete buildings", Proceedings of the 12 World Conference on Earthquake Engineering, Paper No.
0062.
Paulay, T., (2001a), "A redefinition of stiffness of reinforced concrete elements and its implications in
seismic design", Structural Engineering International, Vol.11, No.1, pp. 3641.
Paulay, T., (2001b), "Some design principles relevant to torsional phenomena in ductile buildings",
Journal of Earthquake Engineering, Vol.5, No.3, pp. 273300.
Paulay, T., (2002), "An estimation of displacement limits for ductile systems", Earthquake
Engineering and Structural Dynamics", Vol.31, pp. 583599.
Priestley, M.J.N., Kowalsky, M.J., (1998), "Aspects of drift and ductility capacity of rectangular
structural walls", Bulletin of the New Zealand Society for Earthquake Engineering", Vol.31, No.2, pp.
7385.
Priestley, M.J.N., (1988), "Brief comments on elastic flexibility of reinforced concrete frames and
significance to seismic design", Bulletin of the New Zealand Society for Earthquake Engineering,
Vol.31, No.4, pp.246259.
Tamás PAULAY (1923) PhD, Drs.h.c., professor emeritus at the University of Canterbury,
Christchurch, New Zealand., commenced his studies in 1946 at the Budapest University of
Technology and Economics and graduated in 1953 in New Zealand. After 8 years of structural design
practice, from 1961 till his retirement in 1989, he was teaching at the University of Canterbury where
he also conducted research relevant to the behaviour of reinforced concrete structures subjected to
large earthquakes. He is a past president of the International Association for Earthquake Engineering.