226

SEISMIC DESIGN OF BRIDGES

J.B.S. Huizing*

ABSTRACT:
Seismic Design of Bridges is a comprehensive set of
12 papers presentedin Code and Commentary form. It sets
out the recommendations of the Bridge Committee of the
New Zealand National Society for Earthquake Engineering.
The papers are aimed at experienced designers and
authorities setting design standards. The procedures
recommended are based on the present state of the art
and represent good design practice in accordance with
the latest developments in earthquake engineering as
applied to bridges.

Subjects discussed include: design philosophy, seismic

loading and ductility demand, capacity design principles
and practice for ductile structures, a semi-emperical
approach for structures of limited ductility, ductility
capability and detailing for ductility, structural and
non-structural details, bridge foundations, earth
retaining structures, mechanical energy dissipating
devices, dynamic analysis, structures requiring special
studies and strengthening for seismic loads.

INTRODUCTION: - Existing documents such as the Highway

Bridge Design Brief and NZS 3101 P
1.0 Background be used as a basis for the group's
deliberations, but a completely
Early in 19 78 a working group fresh look should be taken at the
concerned with the seismic design of bridges content of these documents.
was formed on the initiative of and under
the auspices of the New Zealand National - The group * s findings be presented as
Society for Earthquake Engineering. The recommendations plus commentary aimed
first meeting of the working group was at expert bridge designers and author-
held on 13 April 1978. ities setting design standards and not
as a "cook book" or code of practice
The idea behind the formation for designers inexperienced in
of the group was to bring together a bridge design.
number of leading researchers and bridge
designers and to provide them with the - The group's findings be published in
opportunity of discussing and appraising the bulletin and presented at workshops.
the rapid development during the seventies
in bridge research and design. The - A wide range of bridge types and spans
new draft concrete design code, DZ 3101, be examined.
and the experience gained in the use of
the Highway Bridge Design Brief and its - Consideration be given to the importance
associated documents were of course of the highway in an earthquake and to
highly relevant parts of these developments. the economics of achieving seismic
protection.
At the first meeting it was
agreed that the objectives were to make 2.0 GROUP MEMBERSHIP:
recommendations concerning:
The Management Committee in
- Procedures for the seismic design of selecting the group members sought to
bridges and associated structures based cover the various parties most involved
on the present state of the art and in the bridge design and research. The
aimed at practising engineers. membership of the group was as follows:

- Further research work required for Mr J.B.S. Huizing Ministry of Works

resolving remaining uncertainties. (Chairman) and Development

- Dissemination of the recommendations Mr J.F. McGuire Ministry of Works

with explanations to practising bridge (Secretary) and Development
engineers.
Dr J.B. Berrill University of
In addition it was agreed that: Canterbury

Dr R.W.G. Blakeley Ministry of Works

and Development
* Ministry of Works and Development,
Wellington Dr A.J. Carr University of
Canterbury
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
 227
Mr H.E. Chapman Ministry of Works Section Title Sub-Committee
and Development
Small Bridges Fisher, Lanigan,
Mr L.G. Cormack Consulting Engineer Stockwell
Dr F.D. Edmonds Ministry of Works Structural and Lanigan, Preston,
and Development Non-Structural Fisher, Stockwell
Details
Mr R.W. Fisher New Zealand Railways
Earth Retaining Mattewson, Berrill,
Mr P.R. Goldsmith University of Structures Wood
Auckland
10 Dynamic Analysis Wood, Blakeley,
Dr J.M.O. Hughes University of Priestley
Auckland
11 Bridges Requir- Priestley, Stanford,
Dr A.G. Lanigan Consulting Engineer ing Special Carr
Studies
Mr M.B. Matthewson Ministry of Works
and Development 12 Strengthening of McGuire, Fisher,
Bridges for Stanford, North
Mr P.J. North Consulting Engineer Seismic Loads
University of The procedure followed was
Professor R. Park Canterbury for sub-committees to prepare sections
and submit them for scrutiny by the whole
Ministry of Works group. Thus although the proposals
Mr R.L. Preston and Development have been largely prepared by individuals
in the sub-committees, they in general
University of represent the consensus opinion of the
Dr M.J.N. Priestley Canterbury wider group. In the case of Section 9
"Earth Retaining Structures", the group
Ministry of Works did not have the opportunity to scrutinize
Mr P.R. Stanford and Development the proposals in detail although the
approach was approved in principle.
Christchurch City
Mr M.J. Stockwell Council The group met eight times between
April 1978 and April 1980.
Ministry of Works
Dr J.H. Wood and Development 4.0 STATUS:
3.0 MODE OF OPERATION: The recommendations made in
the various reports following this
The group divided the topic introduction represent the considered
into 12 sections and allocated opinion of the group as to good design
responsibility for each section to a practice is accordance with the latest
small sub-committee. The allocation developments in earthquake engineering
was as follows, within each case the as applied to bridges. In several areas
first named acting as initiator and they represent significant developments
editor: in the state of the art. The reports
have no authority as a standard, but
Section Title Sub-Committee they may be used as a guide by designers.
The reports have been written in a form
1 Design Philosophy Stanford, Cormack, suitable for adoption by authorities
Park with responsibilities for preparing codes
2 and design specifications.
Design Loading Berrill, Chapman,
and Ductility Priestley
Demand In accordance with the Society's
objectives it is proposed to organise
workshops for practising designers on
Capacity Design Chapman, North, the application of these recommendations
Principles and Park to specific examples.
Practice
Bridge Found- The Structures Committee of the
Edmonds, Carr, Road Research Unit had a direct interest
ations Goldsmith, North, in the research work which led to the
Wood, Preston, formation of the working group. The
Hughes Structures Committee is vitally concerned
with the dissemination of the research
Detailing for findings and accordingly arranged
Cormack, Chapman, financial support to the Society to aid
Ductility and Park
Ductility publication of the reports and the
Capability running of the workshops.

Mechanical Energy 5.0 CLOSURE:

Dissipating Blakeley, Cormack, It was a privilege to chair such
Devices Stockwell a group of gifted and intelligent engineers.
228

I must admit that at regular intervals

I was somewhat bemused by the learned
discussion. However, most engineers
love to explain their expertise to the
less informed. In the process of these
explanations it proved to be easy for the
group to reach common ground.

The reports speak for themselves

and it only remains for me to thank all
the members of the working group for
their time and effort.

SEISMIC DESIGN OF BRIDGES
SECTION 1
DESIGN PHILOSOPHY
P.R. Stanford*, L.G. Cormack** and R. Park***
1.1 INTRODUCTION
Earthquakes are random occurrences
and a statistical approach involving
consideration of the return period is
called for when choosing the design
intensity. The return period is defined
as the reciprocal of the probability of
the event being equalled or exceeded in
any one year. It is the mean interval
between successive events at that level,
taken over a long time.
The design philosophy recommended
for the provision of earthquake resistance
for bridges, is based on the requirement
that communications should be maintained
to an appropriate standard after all
earthquakes. Ideally, this would be
achieved by ensuring that there is an
acceptable probability of damage occurring
during the life of the structure, at three
broad levels of intensity of earthquake
motion. The three levels are defined
as follows:
(a) In earthquakes with return
periods less than 50% of the life
of the bridge, and thus likely to
occur a number of times during its
life, damage should be minor, and
communications should not be
disrupted.
(b) In earthquakes, with return periods
between 50% and 150% of the life
of the bridge, considerable damage
may occur, but the bridge should
not collapse. The bridge should
be usable by emergency traffic
after simple and rapid repairs,
and should be capable of permanent
repair to an acceptable level of
both vehicle and seismic loading.
The extent to which this is
pursued will be determined by the
controlling authority.
(c) In earthquakes with return periods
greater than 150% of the life of
the bridge, damage may be extensive,
but the bridge should not collapse.
It should be usable by emergency
traffic after temporary repairs,
and may be available for use,
perhaps at a lower level of vehicle
loading, after permanent repairs.
* Senior Design Engineer, Ministry of
Works and Development, Wellington.
** Consulting Engineer, Beca, Carter,
Rollings and Ferner, Auckland
***Professor of Civil Engineering,
University of Canterbury, Christchurch
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
229
As it would be impracticable to
investigate the performance of structures
at all three levels of loading in
every case, it is recommended that only
level (b) should be considered for
routine design. It is expected that if
the design criteria for this level
of loading are satisfied, the performance
at the other levels described will
also be satisfactory.
1.2 ASSESSMENT OF SEISMIC DESIGN LOAD
The seismic design load is based
on an envelope of elastic response of
structures to earthquakes of a specific
return period, in the appropriate
seismic zone. A specific percentage of
critical damping has been assumed in
determing the envelopes.
In most cases it is considered
uneconomic to design a bridge structure
to resist severe earthquakes without
providing means to dissipate the large
amounts of energy which will be imparted
to it by earthquake action. If energy
dissipation can take place, the seismic
design loads will be less than the
elastic response inertia loads. In
consequence, it is recommended that the
design concept should generally be aimed
at providing reliable means of dissipating
energy, either by detailing certain
structural elements (usually piers) to
be ductile, or by providing supplementary
energy dissipating devices.
The level of seismic design load
may be decreased as the design level
of ductility increases (up to a specified
limit), according to an equal displacement
concept for long period structures or
according to a more stringent concept
for short period structures.
1.3 DESIGN LIFE AND EARTHQUAKE RETURN
PERIOD
1.3.1 General case under design earthquake
The design life is not noxmally
a parameter which is specifically stated
in current design criteria, but it is
considered that 100 years is a reasonable
figure to use for this purpose for the
majority of bridges in normal circumstances.
The combination of this life with a 150
year earthquake return period will there-
fore satisfy the criteria of 1.1 (b)
above. This combination is considered
to be appropriate for the following
reasons:
(a) It produces a level of seismic
230
design load similar to that
currently in use. Capacity to
resist this level has been found
from experience to be obtainable
at reasonable cost (5 to 10% of
total cost of the bridge), provided
that the designer considers it at
the outset of the design.
(b) The risk of damage, and the
associated cost of repair during
the life of bridges designed to
this level is considered to be
acceptable.
(c) The combination of 150 year return
period with 100 year design life
implies a probability of
approximately 0.5 that the design
earthquake will be exceeded during
the life of the structure.
1.3.2 General case under extreme earthquake
If structures are designed to
satisfy the requirements of 1.1 (b),
it is expected that they will also satisfy
the requirements of 1.1 (c). During
these larger earthquakes with longer
return periods there will be increased
demand for energy dissipation but it is
expected that this demand could be met
by a well designed structure. This is a
consequence of the following factors:
(a) The actual strengths of earthquake
resisting members are likely to
be greater than assumed for design,
because they will be based on
minimum likely material strengths
and the use of strength reduction
factors. The effects of steel
strain hardening and confinement
of concrete will further increase
the actual strengths.
(b) Due to greater movement between
structural elements, and movement
of foundations in the ground, the
damping is likely to be greater
than assumed.
(c) Under the action of a very large
earthquake, extensive loss of
cover concrete and yielding is
likely to occur, thereby lengthening
the natural period of vibration and
usually reducing structural response.
(d) The displacement ductility factor
which the designer uses to determine
the design load will be the minimum
dependable value. For a well
designed and detailed bridge, a
larger figure is likely to be
achieved in practice.
1.3.3 Bridges of less than average
importance
Notwithstanding the ideal approach
described above, it is acknowledged that
there are many circumstances where it is
not feasible or not economic to provide
a significant amount of ductility, and is
not economic to accommodate the large
inertia forces resulting from the 150 year
event. In such circumstances, the
bridge may be designed for a smaller
return period. The use of 100 or 50
years would imply probabilities of 0.63
and 0.87 respectively that the design
earthquake will be exceeded during the
life of the structure. The justification
for using these reduced periods will be
the importance of the bridge to the
community, and this will be determined by
the controlling authority.
1.3.4 Bridges of greater than average
importance
Conversely, there will be bridges
which may be termed "lifeline structures",
which are essential to communications,
especially after a disaster, and which
require greater than average protection.
If such a bridge is designed for a 250
year return period, this implies a
probability of 0.33 that the design
earthquake will be exceeded during the
life of the structure. These structures
will be identified by criteria laid down
by the controlling authority.
1.3.5 Bridges of other than normal design
life
A major structure may well be
considered as having a design life in
excess of 100 years. Conversely, when
considering a temporary structure or the
addition of seismic protection to an
existing one, the design life may be
considerably less than 100 years. In
both cases, the return period should be
selected to give a probability of exceedance
not greater than that of a bridge of
similar importance designed for 100 years
life.
1.4 THE DESIGN PROCESS
The assessment of the design seismic
load is of course only part of the design
process. Many related problems must be
considered, and the results integrated
to produce an efficient design. There
are three basic skills:
1.4.1 Concept
The first skill is the ability to
conceive in detail the best structural
form. The process will include a number
of specific considerations, all of which
must also be assessed for the effect on
the behaviour of the bridge in resisting
loads other than seismic. The designer
should not forget that even though seismic
considerations may govern many parts of
the structure, its primary function is
always to carry live load.
Specific considerations are:
(a) The bridge may be a framed structure,
a continuous superstructure on
bearings, a simply supported super-
structure , or some combination of
these.
(b) The materials used may affect the
dead load, (and hence the seismic
load) , the flexibility and the

ductility attainable with a
particular configuration.
(c) One or any number of the supports
may be used to resist seismic
forces.
(d) A balance should be maintained
between displacement and strength
requirements. In general, greater
flexibility decreases the response
and hence the seismic loads and
cost of supporting members, but
increases the movements at joints
and hence their cost and the
risk of damage there.
(e) Ductility at any location will
improve the ultimate behaviour,
but there is an arbitrary limit on
the amount which can be used to
lower the level of design seismic
load, and it is possible that the
economic limit may be lower than
that in some cases.
(f) The number and location of energy
dissipating points must be chosen,
whether they are to be plastic
hinges in structural members or
specific energy dissipating
devices. Alternatively, the
structure may be designed to remain
elastic at design earthquake level
in some cases.
231
the structure, and considerable theoretical
knowledge is required. This branch of
earthquake engineering is undergoing
continual development as new techniques
are brought into use. Notwithstanding
the complexity of the analysis, the
choice of a simple structural form is to
be preferred, since the performance of
such a structure will be more reliably
predicted.
1.4.3 Detailing
Finally, the bridge must be
carefully detailed so that all parts of it
behave reliably, and in accordance with
the designer s intentions.
1
This will
involve "capacity" design in which the
primary energy dissipating elements are
chosen and suitably designed and detailed,
and all other structural elements are then
provided with sufficient strength so that
the chosen means of energy dissipation can
be maintained throughout the deformations
which may occur. An equally important
aspect of detailing is that of the move-
ment joints, particularly those which are
intended to accept translational seismic
movements . Adequate clearances between
structural members must be provided for
seismic movements, which are often large
compared with those resulting from length
change effects. The economical design of
joints to cater for both requirements calls
for extreme care.

(g) A choice of the number and location As site conditions may dominate
of movement joints must be made, structural response, the bridge displace-
and whether they are to accommodate ment may well be much larger than that
translation as well as rotation indicated by even the most sophisticated
and whether seismic translation or analysis. Judgement will be required and
only length change effects. high levels of ductility and generous
movement gaps will be often economically
(h) The locations at which damage is justifiable.
to be expected under severe
motions must be identified and the
difficulties involved in repairs
considered.

(i) All of the above must be considered

in relation to earthquake forces in
all three orthogonal directions.

The ability to effectively consider

all the above items is something which can
be gained only through extensive exper-
ience of many types of structure. Simple
adaption of non-seismic designs is
impractical as extensive modifications
(often prohibitive) must always be made.
This is especially true in the case of a
major bridge, where particular care is
necessary to ensure serviceability after
an earthquake. Normally several trials
will be necessary to obtain the best
solution. These trials should be
examined bearing in mind the cost of
providing the earthquake resistance, which
must be commensurate with the value and
importance of the structure.

1.4.2 Analysis

The second skill is the analytical

ability to estimate the response of the
particular structure. In many cases
this is difficult, particularly the assess-
ment of interaction between the soil and
232

SEISMIC DESIGN OF BRIDGES

SECTION 2
DESIGN EARTHQUAKE LOADING AND DUCTILITY DEMAND
J.B. Berrill*, M.J.N. Priestley**, H.E. Chapman***

2.0 NOTATION t = return period, in years

3 = geographic coefficient
= intermediate force coefficient
used in Zone B
^y = seismic displacement of at centre
= intermediate displacement coeffic- of mass relative to ground
A Ay
ient used in Zone B
y = displacement ductility factor
= basic horizontal force coefficient
"Hp
2.1 SEISMIC BASE SHEAR FORCE
= basic horizontal displacement
~HE 2.1.1 Base Shear Force Expression and
= basic horizontal displacement Seismic Zones
"Ay
coefficient
The minimum horizontal seismic base
= correction coefficient for non- shear force H should be derived from the
standard degrees of damping expression

= Young's modulus H = C Z M g (2.1)

E m H

H = seismic base shear force where C.Hy = basic horizontal force

coefficient, and depends on the chosen
I = moment of inertia design value of structure displacement
ductility factor y, the fundamental natural
M = total mass assumed to participate period of the structure, and on the
in horizontal motion seismic zone defined in figure 2.1.
= fundamental natural period of Z = coefficient from Table 2.1,
structure 0

o
= return period coefficient depending on design return period.
M = total mass assumed to participate
= peak vertical acceleration response
in the horizontal degree of freedom. This
= yield stress of steel should normally exclude the mass due to
super-imposed live load.
9 = acceleration due to gravity g = acceleration due to gravity.
I = span length
2.1.2 Basic Force Coefficient C
m = mass per unit length TT

Hy :

= an integer index For Zones A and C, defined in figure

n 2.1, values of C may be obtained directly
ny TT

P = probability from figures 2.2 and 2.4 respectively. For

the transition Zone B, C ^ is the product
Pi = probability of annual exceedance of a geographic coefficient 3 from figure
2.1 and the coefficient
coeffici A from fiqure 2.3.
My
= probability that motion s will be That is, for Zone B
exceeded n times in t years
(2.2)
"Hy
= ground motion with return period
t The design value of the structural
s displacement ductility factor y, is defined
= design life, in years as the ratio of maximum displacement under
the design earthquake to the theoretical
yield displacement, both measured at the
* Senior lecturer in Civil Engineering, centre of mass. Design values of y should
University of Canterbury, Christchurch not exceed six for any structure, unless
** Reader in Civil Engineering, University special studies are carried out to justify
of Canterbury, Christchurch them.
***Ministry of Works and Development,
Wellington In assessing the appropriate value
B U L L E T I N OF T H E N E W Z E A L A N D N A T I O N A L S O C I E T Y FOR E A R T H Q U A K E E N G I N E E R I N G , V O L 13 NO. 3 S E P T E M B E R , 1980
 233

KILOMETRES

FIG. 2.1: SEISMIC ZONES AND GEOGRAPHIC COEFFICIENT B

234

Period ,T (sec)

F I G . 2.2: B A S I C F O R C E C O E F F I C I E N T C H u , Z O N E A

1 , 5
~ ' ZONE B
Coefficient A H U

Peri'od.T (seel

F I G . 2.3: I N T E R M E D I A T E F O R C E C O E F F I C I E N T A H u , Z O N E B

0 0.5 1.0 1.5 2.0 2.5 3.0

Period,T (sec)

F I G . 2.4: B A S I C F O R C E C O E F F I C I E N T C H u , Z O N E C

for y, account should be taken of the
inherent ductility capacity of the materials
adopted, the extent to which ductile response
is assured by the adoption of special
detailing provisions, and the relationship
between structural and member ductility
factor, including effects of foundation and
bearing flexibility.
Calculation of the fundamental natural
period for obtaining the value of C y should
235
where C, = basic displacement coefficient
from figures 2i. 5 or 2 . 7 for
Zones A and C (respectively,
for the chosen value of design
structure displacement ductility
factor y
Z^ = coefficient from Table 2.1
corresponding to the earthquake
return period
H

be based on cracked-section moment of area For Zone B, C. for use in (2.3) is

of piers, as appropriate, and should include Ay
effects of additional flexibility resulting the product of the intermediate displace-
from foundation and bearing deformations. ment coefficient A ^ from Figure 2.6,
and the geographic coefficient 3 from
2.1.3 Return Period Coefficient Z„ figure 2.1. That is, for Zone B, A y

ri
is given by the expression:
The return period chosen for deter-
mining Z from Table 2.1 should be based
H
A.. = 3 A Z (2.4)
A y H
on the design life of the bridge and the y
acceptable risk of occurrence of the
design level earthquake during the design 2.2.3 Relative Ground Displacements
life of the bridge. Between Supports
Attention is drawn to section 10.4
TABLE 2 . 1 - COEFFICIENT Z. where the possibility of relative displace-
ments of piers due to out of phase ground
motions is discussed. This effect should
Return Period (yrs) be considered when span lengths exceed 200 m.
Z
H
2.3 VERTICAL SEISMIC RESPONSE
5 0. 17
2.3.1 General Considerations
10 0.24
20 0. 35 The response of bridge superstructures
to vertical ground motions during seismic
50 0.56 attack should be investigated in the design.
100 0. 80 Bridge superstructures should be designed
to ensure that such response remains within
150 1.00 the elastic range of material behaviour.
250 1. 33
In calculating maximum stresses during
vertical response, neither live load nor
concurrent vertical and horizontal response
2.1.4 Angle of Seismic Attack need be considered.
The design level earthquake should 2.3.2 Vertical Acceleration Response
be considered to act in any direction in
the horizontal plane. However, simult- Peak vertical absolute acceleration
aneous shaking in two orthogonal horizontal response, a , for regular structures may
directions at design intensity need not be taken as 0.67 times the peak horizontal
v

be considered in assessing the strength acceleration response. That is:

required of energy dissipating elements.

2.2 SEISMIC DISPLACEMENTS

a
v = °- 6 7 C
HE H ^
Z
(2.5)

2.2.1 General where C = elastic horizontal force response

uri

HL
coefficient (i.e. C for y 1) from
Consideration should be given to u

displacements induced by response of the Hy

foundation/pier/superstructure system to
ground shaking, and to the consequences Figures 2.2, 2.3 and 2.4 for Zones
of relative ground displacements between A, B and C respectively.
supports.
Z = coefficient from Table 2.1
2.2.2 Displacement Response corresponding to the earthquake
return period.
Where the seismic structural system
can reasonably be simulated as a single g = acceleration due to gravity.
degree-of-freedom oscillator, the maximum In computing vertical accelerations, natural periods
seismic displacement of the centre of mass, of vertical vibration should be used in obtaining
in mm, may be derived from the expression: COMMENTARY - SECTION 2 values of C .
'HE. T

(2.3) C2.1.1 Base Shear Force and Seismic Zones

C
Ay Z
H
The design value of horizontal earth-
quake base shear force depends on the
236

600

F I G . 2.5: B A S I C D I S P L A C E M E N T C ^ M , Z O N E A

600 ~f i i i | i—i—r—i—|—i—i—i—i—|—i—i—r i—r~i—i—I—i i r

Zone B
Coefficient A

F I G . 2.6: I N T E R M E D I A T E D I S P L A C E M E N T C O E F F I C I E N T A . JJ, Z O N E B

seismic zone, the return period of the
earthquake, the fundamental natural period
of the structure and the design value of
ductility.
The seismic zones are based on both
seismicity observed over the short period
of European settlement * ' C2. 2 c 2
geologic and tectonic evidence of earth-
quake occurrence 02.3, C2.4, C2.5, C2.6
The elastic response spectra underlying
the seismic jcoefficients are based
principally on Smith's^ * ^' C2. 2 ^udy 2
of modified Mercalli intensities, on
unpublished work by Matuschka 2 . 7
and on analysesC2.8, C2.9, C2.10, C2.11
of strong-motion data recorded mainly
in North America and Japan. They are
intended to estimate the average response
at alluvial sites. Because of the
random nature of earthquakes and since
the spectral estimates are obtained from
empirical relationships based on sparse
statistical data, there are large
uncertainties associated with the base
shear forces given by equation (2.1).
The uncertainties can be divided into two
types: those arising from scatter about
an expected, or average, spectrum and
uncertainty in the average spectrum itself.
It is felt that uncertainty of the first
type is fairly well described by the
probability distribution underlying Table
2.1 and Figure C2.2. But the mean spectra,
based on much scantier data, may be
substantially in error, possibly by as
much as a factor of two.
Zone B is intended to provide a
smooth transition, approximating relative
risk, between Zones A and B. However,
there are downward steps of up to 15 percent
in the force coefficient C at some periods
in crossing from Zone A to ^ B and from
B to C. These are small compared with
the uncertainties in the absolute values
themselves and do not warrant the use of
a more complicated scaling procedure.
C2.1.2 Seismic Coefficient C
Figures 2.2, 2.3 and 2.4 give elastic
horizontal acceleration response spectra
( C ) , in units of g, estimated for 150
u r
year return period and 5 percent critical
damping, at alluvial sites for Zones A,
B and C respectively, and the corresponding
inelastic spectra for design displacement
ductility values of y = 2 to y = 6. These
curves (CH y have been derived from the
7
elastic response spectrum shape as follows:
(1) For T > 0.7 sec, the equal
displacement principle is applied.
Namely:
C^ = HE C
Hy
y
(2) For T < 0.7 sec, the following
empirical equation is used:
C
"HE T
^Hy
(y-DT
{ +
1}
0. 7
237
Thus, for short period structures the
design curves result in greater design
forces than would be obtained on the basis
of the equal displacement principle, which
is recognized as being non-conservative
for short period structures. At T = 0 ,
the seismic design force is independent
1
a n d of the chosen value of design ductility.
# It is felt that the value of 5 percent
equivalent viscous damping assumed in
figures 2.2 to 2.4 is a reasonable,
s average value for concrete bridges. How-
ever , when response is expected to remain
c
? elastic, or with low ductility demand,
and material damping is expected to be low,
values predicted by this approach may be
nonconservative. Design base shears
for other values of damping may be estimated
by multiplying the value of H obtained
from equation (2.1) by the factor D taken
from Table C2.1. ^
The definitions of yield displacement
and structural ductility factor
TABLE C2.1 FACTOR D^
Percentage of
Critical Damping
2 1.4
5 1.0
10 0.8
are illustrated in figure C2.1, where
point A is the idealized yield point and
point B is the point at which the tensile
reinforcement first reaches yield stress.
In assessing the yield displacement, elastic
stiffness of concrete piers should be based
on the cracked-section moment of inertia,
calculated in accordance with the guidelines below.
TT
Experimental evidence indicates that simple
H y reinforced concrete bridge columns with
fixed bases, detailed in accordance with
the draft N.Z. Concrete Code DZ3101, can
sustain member displacement ductilities
in excess of y = 8. Therefore, such
structures can be designed with confidence
for the lowest value of coefficient C .
TJ
Hy
The value of overall structure ductility
is limited to six to allow for uncertainties
in relationships between structural and
curvature ductility, and to avoid damage
under frequent minor earthquakes.
Where reinforced concrete bridge
columns do not comply with the confining
requirements of the draft N.Z. concrete
(2.7) code, DZ3101, or where foundations or
bearing flexibility increases the ratio
of required local curvature ductility to
required overall structural displacement
ductility, a lower value of structure
ductility should be adopted.
(2.8) The structural stiffness, EI, adopted
in period calculations should be such that
it produces a close estimate of actual
yield displacement of the centre of mass
238

1 1 1 1 ] ! 1 1 1 1 i I I I | I ( I 1 1 I I I I T

Period ,T (sec)

F I G . 2.7: B A S I C D I S P L A C E M E N T C^\i, ZONE C

F I G . C2.2: R E L A T I O N B E T W E E N R E T U R N E D
P E R I O D t D E S I G N L I F E t, A N D
s

P R O B A B I L I T Y O F E X C E E D E N C E P.

of the equivalent simple oscillator.
Consequently the moment of inertia of
bridge piers subject to flexural action
should be based on cracked-section properties,
and contributions to the yield displacement
resulting from foundation and bearing
compliance must be included. It is
recommended that, in the absence of special
studies, the following values should be
used for calculating member stiffnesses:
(a) For members in which primary plastic
hinging is intended to occur (e.g.
pier stems), the EI value is found
from the curvature in the member
at first yield of the tensile
reinforcement. Design aids are
available in CDP 8 1 0 / A - .
C2 13
(b) For members intended to remain
elastic during severe seismic motions
(for example, foundation cylinders),
the EI value is taken as the mean
of the value found from the curvature
in the member at first yield of the
tensile reinforcement and the value
equivalent to the gross uncracked
section of the member.
The response spectra in figures
2. 2 to 2.4 estimate shaking at sites on
deep alluvial soils, typical of most New
Zealand bridge sites. Sites on hard
rock, particularly crystalline basement
rock, may be shaken more strongly at short
periods than comparable alluvial sites.
As a rough guide, the coefficient
may be increased by, say, 20 percent^
for sites on well-cemented sedimentary
rock and by 4 0 percent for crystalline
rock sites, at periods less than 0.4
seconds. At longer periods on hard rock,
lower coefficients may be justified by
special studies, although the possibility
of topographic amplification should also
be considered.
Designers should also be aware of
the possibility of local amplification in
uniform layers of very soft soils such as
unconsolidated esturine sediments. Again,
in these circumstances the state of the
art is not well defined, and special
studies, involving engineering judgement,
are advised.
C2.1.3 Return Period Coefficient, Z„
rl
The coefficient Z„ from Table 2.1
£1
scales the basic force coefficient from
figures 2.2 to 2.4 to produce the design
value of seismic acceleration, as a fraction
of the acceleration of gravity, for
different return periods of seismic attack.
It is expected that in most cases
the bridge owner will specify the design
return period to be used; part 1.3 of
Section 1 : Design Philosophy, makes some
general recommendations. The following
comments are given to help the designer
select a return period when one has not
been prescribed.
The return period t associated
with a given strength of ground
motion s is defined as follows:
239
If p (s) is the probability that
motion s will be equalled or exceeded
during a time interval of one year,
then the return period of motion s is
defined as the inverse of p^ (s),
that is
years (2.9)
Px (s)
Alternatively, it is expected that motion
s will be equalled or exceeded on average
once every t years.
g
To obtain further information about the
likelihood of the ground motion s occurring
during the design life t of the structure
it is necessary to know the distribution
of occurrence times. Statistical data
are sparse, but provided large regions
are considered, they are consistent with
a Poisson arrival process C 2 . 1 4 , C2.15 ^
The assumption that recurrence intervals
are Poisson distributed leads to the
following relationship between return
period, design life and probability of
occurrence:
-1
(1- } (2.10)
fc
s = { 1
where p is the probability that motion s,
with return period t will be equalled g
or exceeded in t years. Expression
(2.10) is plotted in figure C2.2.
It should be noted that the consequence
of adopting a very high probability (say
95 percent) of the design earthquake being
equalled or exceeded during the design
life is that significant damage, requiring
structural repair, can be expected to
result from moderate ground shaking several
times during the life of the bridge. If
such an approach is adopted, careful study
of the full economic and social consequences
should be made. The probability p
(t , t) that the motion with return period
t w i l l be equalled or exceeded exactly n
g
s
times in an interval of t years may be
estimated roughly by the expression
(1 ts } K
t
s
for t s>10
<V t)
n:
(2.11)
which follows from the assumption of a
Poisson distribution of earthquake occurr-
ence .
Finally, it is impossible to estimate
maximum "credible" motions precisely,
but the few existing recordings from
epicentral regions serve as a guide. The
Zone A elastic spectrum (C with u = 1)
multiplied by a factor of 2.25 just
envelops that of the strongest recorded
motions available at present, and it is
considered that these are possible anywhere
in New Zealand. Thus the maximum ''credible"
base shear force for a structure with 5
240
percent
Z = 2.25damping may be (2.1)
in Equation estimated by using
together with
tne appropriate value of C from figure
2.2, for all zones. y
C2.1.4 Direction of Seismic Attack
The design earthquake intensity is
considered to represent the worst ground
shaking along any horizontal axis. It
would therefore be inconsistent to consider
vector addition of concurrent attack in
two orthogonal directions. However,
structural design must be based on the most
disadvantageous direction of peak seismic
attack, and consequences of flexural yield
and displacement in diagonal directions
(rather than simply in longitudinal and
transverse directions) should be considered.
Two examples are listed below:
(1) A group of four foundation cylinders
arranged in a square pattern is
likely to be most sensitive to
horizontal loading along a diagonal
rather than a major axis.
(2) A shear key between superstructure
and supports, designed to resist
transverse relative motion while
allowing sliding longitudinally,
can result in concurrent loading
on the support. The longitudinal
force would then be a function of
the transverse acceleration force
and the appropriate friction
coefficient for the shear key. When
any shear keys located eccentric
to the pier centreline can be
loaded in the above manner, a check
should be made of the torsional
capacity of the pier to resist the
eccentric seismic forces transferred
through those keys.
C2.2.1 Seismic Displacements : General
Considerations
The ductility demand of column
plastic hinges in bridge piers will be
affected by both response displacements
of the centre of mass and by relative
displacements of the ground between piers,
resulting from out~of-phase ground motions.
Where span lengths are large, and column
stiffnesses are high, the increase in
ductility demand from this second cause
can be substantial.
C2.2.2 Computation of Seismic Displacements
Maximum expected displacements of
the centre of mass are important in
designing bearings , abutment clearances ,
and seating details. The maximum dis-
placement is found in similar fashion to
the maximum seismic force, by reference
to figures 2.5 to 2.7 for the basic
displacement coefficient. This is then
modified according to the design return
period, by using the coefficient Z
in Table 2.1.
H
The curves in figures 2.5,
2.6 and 2.7 have been derived directly
from the appropriate curves in figures 2.2,
2.3 and 2.4 observing that for simple
yielding oscillators
T 2
C A
p
= y. C ^ 2— —
y
4
g (2.11)
C2.3.1 Vertical Response ; General
Considerations
In general it is expected that vertical
response of bridge super-structures to
vertical components of seismic ground
motion will be satisfactory. However,
with long span structures, particularly
prestressed concrete bridges where
prestress 'overbalances' dead load, it is
possible that vertical response could cause
superstructure distress or even failure.
Although superstructure seismic forces
for short spans are unlikely to be severe,
large transient variations in reactions
at supports may cause problems in
foundation performance, and possible lift-
off at supports.
Unless special ductility detailing is
incorporated in the super-structure, it is
recommended that maximum actions induced
by the combination of dead load plus
vertical response should not exceed the
following limits:
structural steel and mild steel
reinforcing - 0.9 f
prestressing reinforcing stress
- 0.02% proof stress
concrete strain
- 0.002
C2 .3.2 Computation of Vertical Acceleration
Response
The method adopted for calculating
seismic response in the vertical plane
parallels that outlined for horizontal
response. Equation (2.5) assumes elastic
response with 5 percent of critical damping.
Acceleration responses at other values of
damping may be estimated by multiplying the
value of a^ from equation (2.5) by the
appropriate value of D taken from Table C2.
For structures with regular span lengths
I, approximately constant distributed mass/
unit length m, and stiffness/unit length
EI, the fundamental period of vertical
vibration is approximately given by the
expression:
T v = 0.64£ /g~
2
(2.12)
Equation (2.12) assumes the common situation
in which the vertical stiffness of bearings
is high compared with that of the super-
structure . Where this is not the case,
bearing flexibility should be considered
when calculating the vertical period of
vibration. The mode shape corresponding
to Equation (2.12) is sinusoidal with points
of inflexion at supports. Consequently,
and since vertical response is taken as two
thirds horizontal response, maximum moments
occur at midspan and are approximately given
by the expression:
 241

M = + 0.101 (~ C„_) m g£ 2
(2.13) C2.12 Standards Association of New
max — 3 HE ^
Zealand, "Draft Code of Practice
Reaction changes at each end of a simple for the Design of Concrete
span will be Structures", DZ3101, Standards
2
Assoc. N.Z., Wellington.
R = + -^-3—^ — (2.14) C2. 13 New Zealand Ministry of Works and
IT
Development, "Ductility of Bridges
with Reinforced Concrete Piers",
Care should be exercised in extrapolating N.Z. M.W.D. Publication No. CDP
results of equations (2.12) to (2.14) 870/A, 1975.
to continuous bridges with irregular spans
In this case higher mode response may be C2.14 Cornell, C . A. , "Engineering Seismic
significant, resulting in maximum moments Risk Analysis", Bull. Seis. Soc.
at supports. Am. , Vol. 58, 1968, pp. 1583-1606.

C2.15 Esteva, L., "Seismicity", in C.

C2.4 REFERENCES Lomnitz and E. Rosenblueth (eds) ,
Seismic Risk and Engineering
C2.1 Smith, W.D., "Statistical Likelihood Decisions , Elsevier, Amsterdam, 1976.
of Earthquake Shaking Throughout
New Zealand", Bull. N.Z. Nat. Soc.
Earthquake Engineering Vol. 9, 1976,
pp. 213-221.

C2.2 Smith, W.D., "Earthquake Risk in

New Zealand : Statistical Estimates",
N.Z. J. Geol. and Geophys., Vol. 21,
1978, pp. 313-327.

C2.3 Lensen, G., "Late Quarternary

Tectonic Map of New Zealand", N.Z.G.S.
Misc. Map Series No. 12, D.S.I.R.,
Wellington, 1977.
C2.4 Walcott, R.I., "Present Tectonics
and Late Cenozoic Evolution of New
Zealand", Geophys, J.R. Astr. Soc.,
Vol. 52, 1978, pp. 137-164.

C2.5 Walcott, R.I., "New Zealand Earth-

quakes and Plate Tectonic Theory",
Bull. N.Z. Nat. Soc. Earthquake
Eng., Vol. 12, 1979, pp. 87-93.

C2.6 Suggate, G., "Seismotectonics and

Earthquake Risk in New Zealand",
Proc. 2nd S. Pacific Reg. Conf.
Earthq. Eng., Wellington, 1979.

C2.7 Matuschka, T., Pers. comm. to M.J.N.

Priestley, 1979.

C2.8 Trifunad, M.D. and A.G. Brady,

"On the Correlation of Seismic
Intensity Scales with the Peaks of
Recorded Strong Ground Motions",
Bull. Seism. Soc. Am., Vol. 65,
1975, 139-162.

C2.9 Murphy, J.R. and L.J. O'Brien,

"The Correlation of Peak Ground
Acceleration Amplitude with Seismic
I n t e n s i t y and other Physical Para-
meters" , Bull. Seism. Soc. Am.,
Vol. 67, 1977, 877-915.

C2.10 Mohraz, B., "A Study of Earthquake

Response Spectra for Different
Geological Conditions", Bull.
Seism. Soc. Am. , Vol. 66 , 1976 ,
915-935.

C2.11 Seed, H.B., C. Ugas and J. Lysmer,

"Site-Dependent Spectra for Earth-
quake-Resistant Design", Bull.
Seism. Soc. Am., Vol. 66, 1976, 221-
243.
 242

SEISMIC DESIGN OF BRIDGES

SECTION 3
CAPACITY DESIGN PRINCIPLES AND PRACTICE
H.E. Chapman*, P.J. North**, R. Park***

3.1 PHILOSOPHY: 3.3 DESIGN PROCEDURE :

3.1.1 The capacity design approach may be 3.3.1 It is recommended that for
defined as being a design procedure intended structures in which a plastic mechanism
to ensure that various members of a structural is intended or likely to develop, the
frame form a desired hierarchy of strengths. structural analysis and design procedure
comprise two stages:
3.1.2 The aim is twofold:
(a) Design plastic hinge sections
(a) To ensure that plastic hinges to have the minimum required
intended to develop during strong flexural strengths.
ground motions have a specified
minimum dependable strength. 1
(i) Decide structural form and
This prevents undue damage occurr- choose desired location of
ing during the more frequent plastic hinges to allow a
moderate earthquakes. plastic mechanism to
develop.
(b) To recognise that plastic hinges
developing are likely to possess (ii) Carry out elastic analysis
flexural strengths in excess of under specified loads included
the dependable values .
1
Other in the specified load effect
members in the structure intended combinations.
to remain elastic are designed on 4-
the basis of the plastic hinges (iii) Hence determine minimum
developing their overstrength flexural strengths required
flexural capacities. for plastic hinges. Design
these sections to have
3, 2 DEFINITIONS: ^dependable strengths to
match the requirements.
3, 2.1 The following definitions apply
to sections of a structural member: (b) Design all sections other than
the plastic hinges for shear and
The 'ideal strength is the theoretical
1
flexure. Design plastic hinges
limit strength, based on the section geo- for shear.
metry as detailed and on the nominal min-
imum material strengths. (i) Calculate 'overstrength 1

flexural capacities of
The ^dependable strength^ is plastic hinges as designed
related to the 'ideal strength by the 1
in (a) abdve.
strength reduction factor <J>

* Dependable strength = <j) x ideal strength* ( 3 . 1 )

1 1
(ii) Analyse structure assuming
all plastic hinges to have ^
where cf> is less than 1 . developed their 'overstrength
flexural capacities. Hence
The ^overstrength takes account of
1
determine shear and moment
all the possible factors that may con- capacities required for all
tribute to section strength, such as sections other than the
overstrength reinforcement, increased plastic h i n g e s , and design
reinforcement stress due to strain harden- sections accordingly. Design
ing at large deformations and a concrete plastic hinges for shear.
strength exceeding the specified minimum,
due to age, confinement or other clauses. 3>.3.2 For structures such as abutments
anchored to the approaches by a friction
T
Overstrength* = <J>Q X \ d e a l strength* (3.2) slab, it is recommended that the design
procedure used ensures that the strength
where <£Q , the overstrength factor, of the connection exceeds the anticipated
is greater than 1 . force by which the friction slab will be
mobilised within the approach soils.
* Design Engineer, Ministry of Works and
Alternatively a suitable cyclically ductile
Development, Wellington
connection should be used.
** Consulting Engineer, Hamilton
***Professor of Civil Engineering, University
B.4 DERIVATION OF DESIGN CONDITIONS
of Canterbury, Christchurch
FOR STRUCTURA.L MEMBERS

3-4.1 Bending Moment

3.4.1.1 Plastic hinging intended to occur

B U L L E T I N OF T H E NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980

as a primary energy dissipating mechanism
should have a flexural strength derived
as follows:
The structure should be analysed
elastically under the specified design
load effect combination.
The 'dependable* flexural strength
o f the plastic hinges should be not less
than the bending moments assigned to their
locations from this elastic analysis.
3. 4.1.2 Every member resisting the moments
from plastic hinges (both intended or
p o t e n t i a l ) , should have a flexural strength
derived as follows:
The structure should be analysed
as a plastic mechanism, assuming all
intended or potential plastic hinges to
have developed their overstrength
capacities. The'ideal flexural strength
1
o f the resisting member being designed
should be not less than the bending moment
assigned to the member from this analysis.
3.4.1.3 Every member intended to remain
elastic under design' earthquake conditions,
1
and resisting the moments caused by
frictional forces in sliding bearings,
should have a flexural strength derived
as follows:
The 'ideal' flexural strength
o f the pier members, and the dependable
flexural strength of foundation members
should be not less than the moment induced
in the member by sliding of the bearing
during earthquake motions. An upper
limit to the coefficient of friction of
at least 0.15 should be assumed for
PTFE/stainless steel sliding bearings.
Adequate account should also be taken
o f any additional moments which may be
induced in the member as a result of
earthquake motions along both major axes
o f the structure concurrently. Such
additional moments can arise from friction
o r shear stiffness of devices intended to
prevent horizontal movement in a direction
perpendicular to that being considered.
5.4.1.4 Members resisting the moments
caused by shear forces induced in
elastomeric bearings during earthquake
motions equivalent to the design1 1
earthquake, may be divided into two
categories:
(a) those in which plastic hinging
is intended to occur as part
of the energy dissipating
mechanism;
(b) those intended to remain elastic.
Members in category (a) should
be designed as irAA .1.1. Members in
category (b) should have a flexural
strength derived as follows:
The structure should be analysed
elastically under the combination of
external loading effects specified in
the appropriate specification.
243
(i) If the structure is in the
1
partially-ductile category
1
the increase in forces induced
in the elastic members should
be calculated equivalent to
the horizontal limit displace-
ment appropriate to the design.
Then, the 'ideal* pier member
flexural strength and the
'dependable foundation member
1
flexural strength should be
not less than the moment from
this analysis.
(ii) If the structure is in the
'non-ductile' category, no
increase in force is necessary
beyond the initial elastic
analysis. The 'dependable
member flexural strength should
be not less than the moment
from the elastic analysis.
3.4.2 Shear
3.4.2.1 Every member in which plastic
hinging is intended to occur or could
not potentially occur as part of a primary
energy dissipating mechanism and in which
the shear force able to develop is limited
by formation of plastic h i n g e s , should
have a shear strength derived as follows:
The member should be analysed
assuming all plastic hinges governing
its shear to have developed their over-
strength capacities . The 'ideal' shear
strength of the member should equal or
exceed the shear force from this analysis.
3. 4.2.2 Every member (except for those
designed under3.4.2.1) resisting the moments
from plastic hinges (both intended or
p o t e n t i a l ) , should have a shear strength
derived as follows:
The structure should be analysed
as a plastic mechanism, assuming all
intended or potential plastic hinges
to have developed their overstrength
capacities. The \ldeal shear strength
of the member at any location should
equal or exceed the shear force assigned
to that location from this analysis.
3.4.2.3 Every member intended to remain
elastic under 'design' earthquake
conditions, and resisting the shear
forces caused by friction in sliding
bearings, shall have a shear strength
derived as in3.4.2.1.
3.4.2.4 Members resisting the shear
forces induced in elastomeric bearings
during earthquake motions equivalent to
the 'design' earthquake, may be divided
into two categories:
(a) those in which plastic hinging
is intended to occur as part
of the energy dissipating
mechanism;
(b) those intended to remain elastic.
Every member in category (a)
should be designed as inB.4.2.1.
 244
Every member in category (b)
and part of a structure in the 'partially-
ductile ' category should be designed as
in3.4.2. 1.
Every member in category (b)
and part of a structure in the non-
1
ductile ' category, should have a shear
strength derived as follows:
The structure should be analysed
elastically under the combination of
external loading effects specified in the
appropriate specification. The 'dependable
shear strength of the resisting member
should be not less than the shear force
from this analysis.
3.4.2.5 When there is a relatively greater
degree of uncertainty in the structural
analysis (eg, in foundation piles)
members resisting plastic hinge moments
should be desinged to take account of the
possibility of shear forces exceeding
those derived from the plastic analysis
required in3.4.2.2. Where possible, the
ideal shear strength of such members
T T
should be not less than the shear forces
which would develop if overstrength plastic
hinges developed in the member instead of
at the intended locations. Design
judgement should be exercised, taking into
account the economic effect on the
structure of such provision.
3.4.3 Tension in Connections to Friction
Slabs
3.4.3.1 The tension connection between
a structure and a friction slab intended
to act as its anchorage, should be
designed so that the yield force of the
connection at minimum specified yield
stress is not less than the ideal resistance
of the friction slab.
COMMENTARY:
C3.2.1 Calculation of member strengths
should be based on appropriate Codes of
Practice for the material being considered.
In particular, the following standards
should be used:
New Zealand Standard DZ 3101:1980
"Code of Practice for the Design of Concrete
Structures';
New Zealand Standard NZS 3404:1977
lf
Code for Design of Steel Structures .' 1
This is for use in conjunction with
Australian Standard AS 1250:1975 "SAA Steel
Structures Code? It should be noted that
NZS 3404 and AS 1250 are written for use
in design of buildings and are not intended
for bridge superstructure design. They
d o , however, cover strength design of
members in which plastic hinging may occur.
C3.3. 1 Adoption of the philosophy of
capacity design relates to experimental
evidence that suitably detailed structures
have considerable capacity to undergo
post-elastic displacements, provided they
can do so by flexure rather than by shear.
The object of the design procedure is to
ensure that this is achieved.
In most cases the members
involved in the capacity design procedure
will be the piers, foundation footings
or pilecaps, and the foundation piles
or cylinders. Gravity load requirements
usually lead to superstructures having
adequate strength to prevent seismic
plastic hinging occurring in superstructure
members. Designers should, however, be
aware that this may not always be so.
For example, when a concrete deck slab
is not used it may be necessary to provide
additional strength to prevent plastic
hinging in main longitudinal members under
transverse loading.
C3.3.2 In structures requiring stabilis-
ation under seismic loading, usually in
a longitudinal direction, friction slabs
offer a solution. They may be constructed
to act as a ' f u s e , sliding within the
1
approach filling while anchoring the
abutment, and thereby dissipating seismic
energy. Current knowledge does not,
however, extend to the reliable design of
such energy dissipating mechanisms.
It should also be remembered that
appreciable displacement of the abutment
would be associated with such energy
dissipation and that such a structural
arrangement would also lead to ground
motions being fed into the total structural
mass anchored to the friction slab, up to
the acceleration equivalent to the sliding
force of the slab. It is therefore
advisable to use friction slabs only for
stabilising small bridges or abutments
free to move horizontally relative to the
main structural mass.
Ductile devices offer a method
of connecting a superstructure mass to an
abutment with sufficient strength to
resist service loads, but with the
capability of limiting the inertia force
developing in the connection during
earthquakes (see Section 6 ) .
C3.4 Design conditions for various
structural members can be found in the
clauses noted on figures C3.1 (a) and (b).
C3.4.1 Recommendations for member
strengths and their relative values
necessarily involve applying engineering
judgement. The philosophy adopted for
flexural strengths in this section is as
follows:
(a) For members in which primary
seismic energy dissipation is
intended or likely to occur -
a positive margin is specified
for the resisting member strength
over the plastic hinging member,
to take account of the effect
of large, likely member curvature
ductility demands and consequent
strain hardening effects.
(b) For members in which primary
seismic energy dissipation is not
intended to occur, and for which
calculations are made to quantify
seismic effects, such as effects
of dynamic displacements - a
small margin is specified for
 245

With or without
elostomeric bearings
between piers and
superstructures

Shear - clause 3 • 4 • 2 • I

Moment- clause 3-4 • I • I

HE! — Plastic hinges
Shear - clause 3-4- 2 • I

Moment - clause 3 • 4 • 1-2

Shear-clause 3 - 4 - 2 2 8 3-4-2-S

a) With plastic hinge intended to form in the pier.

FIG. C3-I

Sliding bearings
7 • Elastomeric

Moment - clause 3- 4 • 1-3 Moment - clause

Shear - clause 3- 4 - 2 - 3 Shear - clause

No plastic hinge
intended to form

b) With no plastic hinge intended to form in the pier

FIG. C3- I
246
resisting member strength over
the potential plastic hinging
member. This is on the basis
that curvature ductility demands
should, in any case, be small and
consequent damage of little
significance.
In intermediate cases in which
some seismic energy dissipation is
intended in pier members, but less than
would be the case in 'fully-ductile'
structures, the designer should provide
a larger margin of strength in the
foundation, even if it is not up to the
standard set out in3.4.1.2.
C3.4.1.2 It is not intended that this
clause be applied to a member (e.g. a
pier stem) in which plastic hinging
forms. It is normal practice to carry
the flexural reinforcement required in
the plastic hinge to a point past that to
which the length of the plastic hinge will
extend. No extra provision of flexural
reinforcement is necessary beyond, the
plastic hinge.
C3.4.1.3 A small margin of strength for
the foundation member over the pier it
supports is specified in this case since,
theoretically, plastic hinging cannot
occur in the pier. In practice, many
cases would include the need to provide
such a strength margin due to other
considerations - for example the pier
could be required to act in a ductile
manner transversely. In such a case
clause3.4.1.2 would govern foundation
design.
C3.4.1.4 For the purposes of design,
structure types are defined with
reference to the relationship between the
applied horizontal force and the resultant
displacement (6) of the centre of mass
of the structure.
'Fully-ductile' structures are
those in which a plastic mechanism can
form in the structure. The relationship
is essentially one where, after 'yield',
the resultant displacement increases
without appreciable increase in applied
horizontal force (see Figure C 3 . 2 ) . In
addition, "the relationship must apply for
reversing loads and over at least several
cycles, to ensure hysteretic dissipation
of energy.
'Partially-ductile structures
1
are those where some of the earthquake
resisting elements (eg, piers in flexure)
yield while others (eg, elastomeric
bearings at abutments) remain elastic.
With increasing displacement the applied
force increases, although at a decreasing
rate (see Figure C3. 3) .
The title does not imply that there
is a limit to the ductility available
relative to a fully-ductile' structure,
1
but refers to part of the structure being
ductile and part remaining elastic. As
for 'fully-ductile' structures the
relationship must apply for reversing
loads to ensure hysteretic dissipation
of energy.
'Non-ductile' structures are
those where no earthquake resisting
elements yield and the force/displacement
relationship develops neither a yield
'plateau' nor a hysteretic energy
dissipating capability. The force/
displacement relationship includes elastic
behaviour leading to sudden and irreversible
reduction of load capacity (see Figure
C3.4).
C3.4.2 Design of members to resist
shear forces should be conservative, and
this is reflected in the recommendations
in this section. Because shear failures
are not of a ductile nature nor reversible
in behaviour, it is important that flexural
yield of members be assured before shear
failure occurs. Furthermore, it is
usually not very expensive to provide
added shear strength to members.
In addition to the predictable
shear forces which would develop through
a structure forming a plastic mechanism,
there are several other sources of unpre-
dictable shears or torsional effects
which must be provided for by an adequate
safety margin. These include such
possibilities as torsional ground motions
or structural response, skewed bridge
effects torsion due to out-of-phase
transverse response at adjacent piers, and
concurrent longitudinal and transverse
motion causing friction to be mobilised
in restraints which are intended to
prevent horizontal movement in a direction
perpendicular to that being considered.
Thus the design basis for shear
set out inZA.2 should be regarded as the
least conservative which is acceptable.
In structures where some of the unpre-
dictable effects listed above are likely
to be of consequence, extra shear resist-
ance should be provided.
C3.4.2.5 Dependable analysis and deter-
mination of a heirarchy of structural
member strengths is difficult for some
structures. For example, when piles,
either raked or vertical, pass through
soft soil layers and appreciable water
depths before entering the pilecap, the
structural simulation and dynamic
representation of the structure become
more approximate. Dependable prediction
of the location of the plastic hinges
becomes more difficult. Precautions to
ensure that foundation members can resist
the shear forces arising from a plastic
mechanism inadvertently forming within
the foundation members should be
considered.
C3.4.3.1 Recommendations for calculating
the slidng resistance of friction slabs
are included in Section 4 - Bridge
Foundations.
 a ) Idealised b) Actual

FIG.C3-2 I 'FULLY DUCTILE '

FIG. C 3 - 3 FIG. C 3 - 4
1
PARTIALLY - DUCTILE' ' NON - D U C T I L E '
248

SEISMIC DESIGN OF BRIDGES

SECTION 4 - BRIDGE FOUNDATIONS
F.D. Edmonds*, A.J. Carr*\ P.R. Goldsmith***, P.J. North****, J.H. Wood*****,
R.L. Preston******
4.0 NOTATION soil spring constant; units F/L
k = k, x pile frontal area
s n
= soil undrained shear strength
= k DL h

= basic horizontal seismic coeffic-

ient , see Section 2 = K L H

C (T=0) = basic horizontal seismic

Hvi
K
yy yG' ey' ee
, K K K foundation stiffness
coefficient for period T=0 seconds matrix coefficients,
see Section C4.3.2.1(a)
C (T=0) = 0.52
Hy (Zone A)
length
= 0. 15 (Zone C)
embedded length of pile
D pile width or diameter
cantilever length for stiffness,
EI pile flexural rigidity (pile see Figure C4.2
modulus of elasticity x second
moment of area) cantilever length for moment,
see Figure C4.2
soil modulus of elasticity
M bending moment
F force
constant of horizontal subgrade
H horizontal force reaction (rate of increase in
K with depth); units F/L
h
3

equivalent MM intensities for

1
A' B' C
1 1

the design earthquake in n

h h = d ( K } / d z

seismic Zones A, B, and C

respectively.
coefficient of horizontal sub-
grade reaction (ratio of soil
pressure to lateral pile
deflection); units F/L
standard penetration test resist-
ance (blows per 0.3m)
N, N specific values of N used in the
crit determination of the liquefaction
3

potential of sands, see Equation

4.4
k h = p/y
4
/EI * Relative stiffness factor
modulus of horizontal subgrade / for piles in uniform
reaction (ratio of soil pressure V cohesive soils, see Section
to pile lateral diametric strain) C4.3.1 (a)
units F/L 2

R' EI Relative stiffness factor

P n^ for piles in uniform
Y/D cohesionless soils, see
Section C4.3.1 (a)
2
P
* Design Engineer, Ministry of Works and soil pressure; units F/L
W
Development, Wellington
** Senior lecturer in Civil Engineering, weight
University of Canterbury, Christchurch y
*** Consulting Engineer, Auckland pile lateral deflection
z
**** Consulting Engineer, Hamilton
***** Laboratory Services Engineer, MWD, depth from ground line
z depth of water table below
Central Laboratories, Lower Hutt.
******Assistant Chief Designing Engineer, ground surface
MWD, Wellington. seismic zone B geographic
coefficient, see Section 2

rotation
strength reduction factor

effective soil internal angle

of friction

structural ductility factor

coefficient of friction

4.1 FOUNDATION/STRUCTURE INTERACTION
4.1.1 In assessing the response of a
bridge structure to earthquake excitation,
the foundation flexibility resulting from
the interaction between the bridge
foundation and the soil should be taken
into account.
4.1.2 In the design of minor bridge
structures or as a first approximation
in the design of a major structure, the
effect of foundation flexibility on the
design horizontal seismic coefficient may
be disregarded if the rigid base period
(the natural period of the structure
assuming fixity at the ground line) is
greater than 0.25 second.
4.1.3 In assessing the seismic loading
and the curvature ductility demand at each
bridge pier, account should be taken of
differences in foundation flexibility at
the various bridge piers and abutments.
4.1.4 The effect of foundation flexibility
should be taken into account in assessing
the seismic displacements at the bridge
piers and abutments.
4.2 FOUNDATION DESIGN
4.2.1 Because of difficulties associated
with the repair of foundation damage
resulting from a severe earthquake, it is
desirable to design the bridge foundations
to remain elastic below the ground line
at the design earthquake loading. In
the case of flooded foundations it is
similarly desirable to design the pier
stem below the mean high water line to
remain elastic.
4.2.2 In general, bridge foundations should
be designed to resist the imposed seismic
loading in accordance with the principles
of capacity design as outlined in Section
3.
4.2.3 Where yielding of the bridge
foundations at the design earthquake
loading is unavoidable, the foundations
should be designed to behave in a ductile
manner by using the detailing guidelines
given in Section 5.
4. 3 PILE OR CYLINDER FOUNDATIONS
4.3.1 General Considerations
4.3.1.1 In assessing the 'effective ground
line' (the level below which soil inter-
action can be expected to influence the
seismic response of a bridge structure)
the following considerations should apply:
(a) Cohesive soils. The effective
ground line should be assumed at a
depth of 1.5 to 2 pile diameters
below the actual ground line.
(b) Cohesionless soils. Where the
surface materials have a standard
penetration resistance, N < 2 ,
the effective ground line should
be assumed at a depth underlying
such materials.
249
4.3.1.2 The lateral resistance from
friction along the underside of a pile
cap should be disregarded for design
purposes as it may be effectively
eliminated by soil settlement during an
earthquake.
4.3.1.3 In cohesionless soils, additional
lateral restraint from passive pressure
mobilised against an embedded pile cap
may be used in the design provided scour
or excavation around the pile cap is not
likely to render it ineffective.
4.3.2 Analysis of Pile Foundations
4.3.2.1 A number of methods are available
to analyse pile foundations.
These include:
(a) Approximate methods using soil
stiffness parameters which are
assumed to be independent of
the soil strains:
(i) Equivalent cantilever methods
(ii) Analytical solutions based on a
soil/pile system analogous to a
beam on an elastic foundation.
(iii) Analytical solutions based on
the assumption that the soil acts
as an elastic half space.
(iv) Computer solutions using soil
springs attached at discrete
intervals along the embedded
length of the pile to model soil/
pile interaction.
(b) Methods using soil load/deflection
characteristics which are compatible
with the soil strains mobilised
by the applied loading. Such
methods normally employ soil springs
as in (iv) above, the spring stiff-
ness being adjusted to match the
secant stiffness corresponding
to the pile deflection at each
soil spring location.
(c) Refined soil foundation interaction
models. Refined models which may
be more appropriate for use in
time-history dynamic analyses,
using soil springs or a finite
element mesh to represent the
soil.
4.3.2.2 Where preliminary analysis indicates
that the seismic response is significantly
affected by soil stiffness, the design
should be based on analysis in which the
soil stiffness parameters are compatible
with the pile deflections produced by the
design loading.
4.3.2.3 Rational analysis should be used
to determine the manner in which raked
piles resist lateral loading.
4.3.2.4 The effect of interaction between
piles in groups on the foundation stiffness
should be taken into account. In the
absence of more specific information, the
250
effective modulus of horizontal subgrade
reaction for each pile may be assumed to
vary linearly from 25% of the value
for a single pile when spacing xn the
direction of load is 3 pile diameters to
the value for a single pile when the
spacing is 8 pile diameters.
4.3.2.5 Unless otherwise substantiated,
the equivalent viscous damping of the
bridge soil/structure system should be
taken to be 5% of critical damping. In
a time-history analysis in which soil
damping is taken into account by modelling
the soil as an inelastic material or as
an inelastic material with viscous damping,
the damping in the structural system
should be taken as 3% of critical, see
Section 10.
4.3.3 Design Loading for Pile Foundations
4.3.3.1 Pile foundations designed in
accordance with the principles of capacity
design. The piles should be designed
to resist the loading imposed by the
formation of plastic hinges in the bridge
piers, in accordance with the requirements
of Section 3. In the case of rigid
1 1
piles (L'/R or L / R < 4 ) , the lateral
1 F
load resistance of the pile foundation
as governed by the soil in which the piles
are embedded, should also comply with:
0 x (lateral load resistance of the
foundation*)
5* (loading inposed by * overstrength'
plastic hinging in the piers) (4.1)
where 0 is the appropriate soil strength
reduction factor listed in Table 4.1.
4.3.3.2 Bridges with limited ductility
where yielding of the piles is expected
to occur under the design seismic loading.
The piles should be designed to resist
the seismic loading (corresponding to
the appropriate ductility factor y listed
in Section 7.5.2) in accordance with the
requirements of Section 3. In the case
of 'rigid' piles (L'/R or L 7 R ' < 4) , the
lateral load resistance of the pile
foundations, as governed by the soil in
which they are embedded, should also
comply with:
0 x (lateral load resistance of the
foundation *)
> (design seismic loading) (4.2)
where 0 is the appropriate soil strength
reduction factor listed in Table 4.1.
4.4 SPREAD FOOTINGS
For spread footings, conventional
design techniques should be used. Where
the overturning moments exceed the
economical limits for conventional capacity
designed footings it may be practical to
consider a pier design in which rocking of
the pier and its footing on the soil becomes
^Methods by which the lateral load resist-
ance of single and group pile foundations
may be evaluated are referenced in the
Commentary.
the lateral load limiting mechanism.
4.4.1 Rocking Foundations
4.4.1.1 Bridge piers may be designed to
rock on their foundations at a design
seismic coefficient corresponding to y = 3
when it is uneconomic to provide sufficient
stability to prevent rocking. However
the advantages of a rocking response
indicated by recent research are recognised
and may be taken advantage of at a lower
level of response, corresponding to
3 < y < 6, provided special studies are
carried out to ensure satisfactory behaviour.
4.4.1.2 Piers on rocking foundations are
not required to meet the special seismic
requirements for reinforcing outlined
in Section 5.
4.4.1.3 The design vertical load on a
bridge pier intended to rock under the
design seismic loading should be taken
as 0.8 times the dead load and overstrength
contributions from superstructure or other
members adjacent to the pier which may
yield during rocking of the pier should
be taken into account.
4.4.1.4 To protect against excessive
plastic deformation of the soil imposed
by rocking, which may result in
misalignment of the pier after an earth-
quake , it is suggested that the bearing
pressure on the soil under the design
lateral loading should not exceed the
ideal bearing pressure divided by 1.8
as recommended in Reference 4.2.
4.5 FRICTION SLABS
4.5.1 Friction slabs may be used to
provide seismic anchorage at bridge
abutments in accordance with the capacity
design principles outlined in Section 3.
The dependable resistance provided by the
friction slab should be calculated as
follows:
H = 0 W (y' - C (T=0))
Hy (4.3)
where W is the dead weight of the friction
slab and the overlying fill, C (T=0) is
R
the basic horizontal seismic ^
design coefficient for period T = 0 seconds,
y' is the coefficient of friction between
the friction slab and the cohesionless
material against which it has been cast
and 0 is an appropriate strength reduction
factor.
4.5.2 Friction slabs should be founded
on cohesionless material. Where friction
slabs are to be used in a cohesive material,
a layer of coarse granular material of
at least 100 mm thickness should be
placed beneath the friction slab.
4.5.3 Friction slabs should not be
used to provide seismic anchorage in an
embankment which is likely to fail under
earthquake loading.
4.6 SOIL PARAMETERS
4.6.1 Analyses should be carried out
 251

for a range of soil parameters to test for A check on the lateral load
sensitivity and to provide a design resistance of pile or cylinder foundations,
envelope for all likely soil conditions at as governed by the soil in which the piles
a particular site. Suggested soil stiff- are embedded, should be made if:
ness parameters for use in preliminary
seismic analyses of pile foundations are
given in Table 4.2. L ' / R '< 4 Cohesionless soils
4.6.1.1 Variation of Soil Stiffness with 5
Depth. In an analysis in which where R' = / —
the soil stiffness parameters are assumed n
h
to be independent of the mobilised soil
strains, see Section 4.3.2.1 (a) the
following may be assumed to apply: or L'/R < 4 Cohesive soils
(a) Uniform cohesive soil. Depending 4
on the method of analysis used,
where R =
(i) the soil modulus of elasticity
E may be assumed to be constant
s J

or
(ii)the modulus of horizontal subgrade
reaction may be assumed to be
constant with depth and to be
independent of the pile width or
diameter.
(b) Uniform cohesionless soil. The
modulus of horizontal subgrade
reaction may be assumed to
increase linearly with depth and
be independent of pile width or
diameter.

4.6.2 Soil stiffness parameters more

appropriate to a given bridge site may be
determined from the soil stress/strain
characteristics or the soil resistance/
deflection characteristics obtained from
laboratory testing of samples or from
insitu testing.

4.6.3 The effective reduction in the

soil sti ffness due to cyclic loading under
earthquake conditions should be accounted
for by using stiffness parameters equal
to 70% of their static values. (Note
that the suggested values given in Table
4.2 incorporate this reduction).

4.7 LIQUEFACTION

4.7.1 The liquefaction potential of

soils comprising loose saturated sand and/
or coarse silts should be investigated.
The recommended method of assessing
liquefaction potential is outlined in the
commentary.

4.7.2 Important or major bridge structures

should not be located at sites where the
potential for liquefaction exists. In
some circumstances where the potential for
liquefaction is very localised, some
remedial measures can be taken as discussed
in the Commentary.

TABLE 4.1 Strength Reduction Factors for the Evaluation of the Dependable' Lateral
1

Load Resistance of Pile or Cylinder Foundations.

Data Available Strength Reduction Factor* *From Table 3,

Reference 4.1
Test loaded piles, plate bearing 0.70 - 0.90
tests or insitu pressuremeter tests
Measured soil parameters 0.60 - 0.70
Visually assessed soil parameters
;
:
from bore logs 0.50 - 0.60
252
T A B L E 4.2 Suggested Soil Stiffness Parameters for use in a Preliminary
Seismic Analysis

SITE DATE DESIGN PARAMETERS

SOIL TYPE E
N C r n
h <
kN/m ) 3 K
h
s
(blows/0.3m) u (degrees) Submerged (kN/m^)
Dry (kN/m )2

(kPa)
Cohesionless (dense 30 - 50 45° 15xl0 3
9xl0 3

soils (
(loose 4-10 30° 2xl0 3
lxlO 3

„ , . [hard 20 - 60 150-250 18xl0 25xl0

3 3

Cohesive . 8-15 50-100 6xl0 8xl0

3 3

2 - 4 15- 30 1.5x103 2xl0 3

s o l l s
jsoft

SITE DATA: at the pier hinge, an increase in the

N =
C =
Standard penetration test
resistance (blows/0.3m
penetration)
Soil undrained shear strength
yield displacement due to an increase in
foundation flexibility will reduce the
available overall structural ductility.
For a rigid base period greater than that
corresponding to peak seismic response,
the increase in C due to the reduction
up 0

m the available structural ductility is

0 1
= Effective soil internal angle almost matched by the reduction in C
of friction due to the increase in period. It ^
can be shown from the design response
DESIGN PARAMETERS spectra (Section 2) that for each
Seismic Zone, C for a flexible
0

Hp
(a) Cohesionless soils
foundation is less than about 1.25 times
_ constant of horizontal the rigid base value provided that the
rigid base period is greater than 0.25
n

subgrade reaction
second.
= d(K )/dzh C4.1.3 Unequal foundation stiffness in
an otherwise symmetrical structure will
= d(k,D)/dz influence the seismic shear load and
curvature ductility demand at each bridge
pier. The piers with the stiffer
(b) Cohesive soils foundation will resist a relatively greater
proportion of the lateral load and have a
(i) K, = modulus of horizontal greater curvature ductility demand placed
subgrade reaction upon them.
= k, D
n C4.1.4 Although C ^ is relatively
or (ii) E = soil modulus of insensitive to foundation flexibility,
elasticity increased flexibility of the foundations wil
significantly increase the seismic deflect-
ions .
COMMENTARY: C4.2.1 The bridge designer would normally
be expected to make a preliminary cost
C4.1.1 To assess the seismic loading and assessment of foundation repair following
displacement of a bridge structure it is a severe earthquake as compared to the
first necessary to establish the stiffness additional cost involved in designing for
(and in more rigorous analyses, the an elastic foundation, taking into account
damping) that the structure derives from the design life of the bridge.
its foundations.
C4.3.1 In estimating the foundation
C4.1. 2 A ductile structure of given stiffness, only that part of the soil that
natural period of vibration and flexural can be relied upon to provide lateral
yield strength will respond to the design restraint under seismic conditions should
earthquake with a certain structural be taken into account. A low estimate of
ductility demand. An increase in the likely foundation stiffness will in
foundation flexibility will increase the general result in a conservative design,
natural period of vibration and increase i.e. an overestimate for the seismic
the yield displacement of the bridge displacements and, as discussed in
structure. For a given curvature ductility Section C4.1.2 , an overestimate for the
design seismic loading.

The recommendation concerning
cohesive soils is made because of the
potential for pile/soil separation at
the pile top under cyclic loading and
is based on the findings of Davisson and
Prakash • and Matlock - .
4 3 4
Cohesionless
4
soils have a greater ability to flow into
the separation gap. The recommendation
concerning cohesionless soils with N< 2
is based on the suggestion of ACI Committee
3364.5 m Where great depth of such
cohesionless material exist, special
consideration of their stiffness and
strength may be appropriate though such
soils would not normally be suitable for
bridge foundations in seismic areas
because of their potential for lique-
faction , see Section 4.7.
Shaking table tests by K u b o • 4
have confirmed that the passive pressure
acting on the vertical face of embedded
pile caps in cohesionless soils
significantly reduces the moments induced
in the foundation piles.
C4.3.2 The use of a more refined soil
foundation interaction model for pile
foundations will not necessarily lead
to a more reliable prediction of found-
ation behaviour as the accuracy of the
prediction will depend as much on
the reliability of the soil data as upon
the refinement of the model. Confidence
in the soil data implies knowledge of the
following:
Modification to the undisturbed
characteristics of the soil caused
by the change in the stress state of
the soil during and subsequent to the
installation of the foundation.
Time dependent changes in the soil
properties depending on the number
of cycles of loading and the amplitude
of each cycle.
The development of methods to solve the
soil/foundation interaction problem have
been the subject of considerable research
effort in recent years, as discussed by
Seed et a l • and Berger and P y k e - .
4 7 4 8
However because of the complexity of the
physical problem advances that have been
made in predicting soil parameters
suitable for use in seismic soil foundation
interaction analyses from standard soil
tests have not kept pace with advances
made in the analytical techniques. Caution
should thus be used in relying too heavily
on the results of soil testing, however
extensive, as relating the test results
to the physical conditions in the soil
around a pile foundation under seismic
loading will inevitably involve consider-
able judgement.
C4.3.2.1 (a) For preliminary seismic
analyses of a bridge structure or for
the purpose of design where it can be
shown that the seismic response is relat-
ively insensitive to changes in foundation
flexibility, use may be made of the
simplifying design assumption that the
soil stiffness properties are independent
of the strain level. For seismic design
253
purposes, the elastic soil stiffnesses
used in such analysis should relate to
the secant stiffness of the soil resistance/
deflection relationship at approximately
one half of the ultimate lateral soil
resistance. As the lateral response
of a pile foundation is largely governed
by the interaction of the piles over the
top 8 to 10 diameters, the representative
soil stiffness in relatively uniform
soils may be based on the soil character-
istics at a depth 4D from the ground line.
A suggested range of soil stiffness
parameters for use in preliminary seismic
analyses of pile foundations is given
in Table 4.2.
For the purpose of analysis, an
artificial division is often made between
6
the structure above the ground line and
the bridge foundation. Using dummy
members whose load/deflection relationship
at the ground line approximates that of
each foundation group and the surrounding
soil (see fig C4.1(a)) the seismic
response of the bridge structure can be
obtained from the design response spectra
(Section 2) or by means of time-history
dynamic techniques (Section 10).
The peak pile bending moments,
or if required the distribution of bending
moment along the length of the piles,
corresponding to the design seismic
loading can be subsequently obtained from
standard solutions available in the
literature. In the case of a Winkler
soil spring model (see fig. C4.1 (b)),
the pile bending moments can be obtained
directly by applying the design seismic
loading.
(i) Equivalent cantilever methods.
This method, which is based on beam on
elastic foundation theory, may be used
as a rough approximation for free head
piles in uniform cohesive soils, where
the soil stiffness can be reasonably
represented by a constant modulus of
horizontal subgrade reaction, or in
uniform cohesionless soils, where the
modulus of horizontal subgrade reaction
can be assumed to vary linearly with
depth. Details of this method, which
involve the evaluation of the relative
stiffness factors:
4
R =y EI /
(for cohesive soils) (C4 .1)
K
h
5
or R = /EI
!
(for cohesionless soils)
V H7 (C4.2)
are given in fig C4.2. Note that differ-
ent cantilever lengths are used depending
on whether foundation stiffness (for the
purpose of evaluating superstructure
displacement) or peak pile moment is to
be modelled. The cantilever lengths for
foundation stiffness are based on
Reference 4.9 and are approximations for
a range of ground line moment to shear
ratios . More stringent elastic theory • 4
 254

I H' - Kyy y e
K
X
y
- ey
K K
eeJ e
///}///

(a) Methods by which foundation stiffness can be

represented by a dummy member
-7777 rrr"
Actual System

(b) Winkler soil spring representation

of pile / soil interaction

FIG. C4 I METHODS OF REPRESENTING PILE FOUNDATION STIFFNESS

Flexural rigidity of pile = E I

Diameter of pile = D

Equivalent Cantilever for Foundation Equivalent Cantilever for Pile

Bending moment Deflection - Stiffness Moment

Deflection at level of applied load Moment at fixed base of

Actual System matches actual system cantilever matches peak pile
moment of actual system

Cohesive Soil EI
R = L s = I-4R L m = 0-44'R
(Constant K )
h

Cohesionless Soil EI
8R' L m = 0-78R'
(Constant n^ )

FIG.C4-2 EQUIVALENT CANTILEVER METHOD


indicates that these cantilever lengths
will vary to some extent with the ground
line moment to shear ratio. The
cantilever lengths for pile moment are
from Reference 4.9 for the case of applied
shear at the ground line and are conser-
vative for the more general case of shear
and moment at the ground line.
(ii) Analytical solutions based on
elastic foundation theory.
Pile foundations can be generally cate-
gorised as rigid' or 'flexible' in the
1
following way:
L'/R or L /R 8 8
< 4
L'/R or L'/R* > 4
where L is the embedded length of the pile.
1
Beam on elastic foundation theory indicates
that the solution for L'/R or L'/R' > 4
is essentially the same as the solution
for L'/R equal to infinity. Graphical
solutions for deflection and moment along
the length of flexible piles are presented
by Davisson • . These solutions cover
4 9
a range of pile head fixity (from free
through to fixed). Note that the majority
of practically proportioned piles will be
in the flexible category.
(iii) Analytical solutions based on
the assumption that the soil
acts as an elastic half space.
Many of the available structural analysis
computer programs, e.g. STRUDL (reference
4.11), have the facility to accept an input
stiffness matrix in place of input stiffness
parameters. Foundation stiffness matrices
of the form:
H pile displacement corresponding to the
K K
M e
which relate ground line shear H and
moment M to ground line deflection y and
rotation 0 can be derived from the work
of Poulos •12 y be obtained for free
4
o r m a
head piles from the following equations:
K. 16 EI
yy ^3
_ K
= 3.2
where EI is the flexural rigidity of the
piles and (E /E) is the ratio of the soil
g
to pile modulus of elasticity. The above
equations are empirical approximations
provided by Blaney et a l -13 to a
parametric study conducted by Novak •^
and are suitable for use in preliminary
analyses where more detailed study is not
255
required. Graphical solutions for peak
pile moment in terms of the design ground
line bending moment and shear force are
provided by Poulos •12.4
The assumption of a constant soil
elastic modulus normally applies to
uniform cohesive soils only and should
not be used for cohesionless soils or
for layered cohesive soils.
(iv) Soil spring models. Computer
models which use soil springs attached
at discrete intervals along the length
of the pile are similar in theory to
(rigid piles) (C4.3) analytical solutions based on a beam on
an elastic foundation. However the
(flexible piles) (C4.4) computer models enable the theory to be
applied to layered soils where the soil
stiffness is not characterised by constant
or linear variation with depth. Margason
and Holloway •15 point to the importance
4
of accounting for layering effects as
rapid changes in soil stiffness with
depth can induce local bending moment
maxima in the piles.
C4.3.2.1 (b) Account of the amplitude
dependent nature of soil stiffness can
most readily be achieved by using a soil
spring computer model in which the soil
stiffness associated with each spring is
appropriate to the peak seismic pile
displacement at each particular soil
spring location. Each soil spring
constant
k s = k.h DL (C4.9)
should be based on the secant stiffness
of the soil p/y curve at the appropriate
n y design loading, see Fig C4.3. Methods
by which the soil p/y curves (the soil
x (C4.5) resistance/deflection characteristics)
can be obtained from laboratory or in-
situ tests are referenced in Section
C4.6.2. The analysis will normally
require an initial estimate of the
deflected pile shape followed by a number
of iterations to achieve compatibility
between the initially assumed displace-
ments and the computed displacements at
each soil spring location. The number
of iterations will depend on the precision
0. 75 (C4.6) with which the solution is required.
In general the computed pile displacements
are more sensitive than pile bending
moments to changes in soil stiffness.
-4.8 EI 0.50 (C4.7) C4.3.2.1 (c) Where it is warranted,
^2 e.g. for use in a time-history analysis,
a more refined model may be used to
represent soil foundation interaction.
EI 0.25 (C4.8) This may be done using a computer model
incorporating inelastic soil springs
D or elastic soil springs with dashpots
to model hysteretic soil damping and
incorporating lumped masses along the
embedded depth of the pile to represent
soil inertia effects. Alternatively a
finite element mesh can be used to
represent the soil, For a time-history
4
analysis, equal seismic excitation should
4 4
be input at each of the soil spring
support points or in the case of a finite
element mesh, at an appropriate soil layer
 H (Design seismic loading )
\ "
\ Family of p / y curves far a specified
\ pile diameter
\
The p/y curves relate the soil resistance
at a point to the pile deflection at the
same point.
The Winkler soil spring constants are
obtained from the secant s t i f f n e s s
of the p / y curve at each soil spring
location ;

k = k. x frontal area
s h
= k, x D x L

Pile diameter = D

FIG.C4-3 DETERMINATION OF WINKLER SOIL SPRING CONSTANTS FROM SOIL P/Y CURVES

boundary at or below the base of the pile.
A comparison of these two methods is
given in Reference 4.8.
C4.3.2.3 Raked piles resist the seismic
loading imposed by the bridge superstruct-
ure partly by axial load and partly by
bending. As shown by Patterson-Kane •^,
4
it is not sufficiently accurate to assume
that raked piles resist the lateral load
purely by axial load, even in situations
where the line of action of the horizontal
load coincides with the intersection of
the axes of the raked piles.
C4.3.2.4 The recommendation to account
for the effective reduction in soil
stiffness resulting from interaction
between piles in a group is based on the
work of P r a k a s h • and Davisson and
4 17
Salley •18.
4
They found the interaction
existed between piles at spacings of up
to 8 pile diameters in the direction of
load and 3 pile diameters in the direction
normal to the load. Recent research
at Auckland University by Hughes et a l * ^
4
indicates that the 8 pile diameter spacing
is non-conservative and that interaction
in cohesionless soils occurs at spacings
of up to 13.5D in the direction of the
loading. This work has also investigated
load sharing between piles in a group and
has indicated that a disproportionate
share of the lateral load is resisted by
the front piles.
C4.3.2.5 Damping in a dynamically loaded
soil/structure system occurs as energy
is dissipated in a hysteretic manner in
the structure and the soil and from energy
radiated from the structure into the soil.
The concept of equivalent viscous damping
to represent the overall effect of damping
in the system is used for mathematical
convenience rather than on the basis of
its correctness.
Traditionally the damping ratio
for bridge structures under design
seismic excitation has been taken as 5%
of critical though it has been thought
that this value would depend to some
extent on the soil type. Present
understanding of the relationship between
an overall viscous damping factor for the
system and the actual damping mechanisms
does not permit any rational modification
to this assumption.
C4.3.3 Piles resist applied lateral
load by the combined action of their
inherent structural strength and
rigidity and the resistance mobilised by
the surrounding soil as the pile deforms.
In general terms, the lateral load resist-
ance of a pile or group of piles is governed
by either the flexural strength of the
piles (in the case of 'flexible' piles)
or by the soil strength (which may govern
in some circumstances for 'rigid' piles).
Thus in addition to designing piles to
have sufficient strength to withstand the
imposed bending moments and shear forces
in accordance with the requirements of
Section 3, a check on the overall lateral
load resistance as governed by the soil in
which the piles are embedded should also
257
be made in the case of 'rigid piles
1
(L'/R or L'/R' < 4) where soil strength
considerations may apply. Note that
for the case of pile foundations designed
in accordance with the principles of
capacity design, the dependable' rather
1
than the 'ideal lateral load resistance
1
is matched to the loading imposed by
'overstrength plastic hingeformation in
1
the piers, in contrast to the strength
requirements of the pile foundation covered
in Section 3.4.1.2. Matching the
'dependable' lateral load resistance to
the design seismic loading for the found-
ation piles has been recommended for both
ductile design and for the design of
bridges of limited ductility as the 0
values listed in Table 4.1 relate to the
certainty with which the soil strength
parameters can be estimated rather than
to an assessment of the dependable
strength of the soil.
Methods by which the lateral load
resistance for single piles (Kuthy et al •20
4
and for groups of piles (Pidgeon and Toan •2 4
may be evaluated are readily available.
Reference 4.21 gives details of a
generalised computer program which may
be used for raked or vertical pile groups
and can also be used for single piles.
These methods pertain to static loading
but may be used to approximate the lateral
load resistance of pile foundations under
seismic conditions.
C4.4 An outline of the different design
considerations for spread footings and
rocking foundations is given in Reference
4.2.
C4.4.1.1 It is generally accepted that
bridge foundations need not be designed
for forces larger than those corresponding
to y = 3. In some circumstances (e.g.
in the design of the footings of slab
piers subjected to lateral loading in
the transverse bridge direction) it may
be difficult or impossible to design a
spread footing to resist the pier moment
capacity. Should the pier moment capacity
exceed the seismic overturning moment
corresponding to y = 3 and the natural
period of vibration (prior to rocking) ,
rocking of the bridge pier and its
footing on the soil may be assumed to
occur.
Because of the complete absence of
experience with rocking piers in earth-
quakes , the design of a rocking foundation
for a load level less than that correspondin
to u = 3 should be based on special studies,
including appropriate dynamic analyses • 22 ^
4
to verify the behaviour of the rocking
system. The lower limit on the load
at which rocking may be permitted to
commence should not be less than that
corresponding to y = 6.
C4.4.1.2 A rocking pier may be considered
sufficiently protected against overload
and hence failure if it possesses the
ideal strength to resist the seismic
forces corresponding to the chosen
ductility factor u. Piers on rocking
foundations are thus exempted from the
258
special seismic requirements for rein-
forcing. However they should be detailed
for limited ductility to ensure some
protection in the event that the base does
not rock until a load greater than that
corresponding to y = 3 is reached.
C4.4.1.3 In assessing the size require-
ments for the rocking foundation pad, a
design vertical load of 0.8 times the dead
load (to account for the effect of
vertical acceleration under earthquake
conditions) should be used to minimise
the pier restoring moment while over-
strength contributions of any yielding
connections should be taken into account
to maximise the overturning moment. In
addition, a thorough analysis should be
carried out to determine the ductility
demands on components of the bridge other
than the rocking pier to ensure that these
do not exceed the demands implied by the
y factors appropriate to those components.
This implies a full assessment of the
performance of both structural and non-
structural components of the bridge as a
consequence of the vertical and horizontal
movements associated with the rocking
motion of the piers. Spans which extend
between rocking and non-rocking piers or
abutments must be detailed to preserve
their integrity for carrying the intended
vertical load.
C4.5.1 The derivation of Equation 4.3
which aims to give a conservative estimate
of the dependable seismic resistance
provided by friction slabs is outlined
in Fig C4.4. The implication is that
the friction slab will not be capable
of providing resistance to seismic forces
in the bridge superstructure if y ' ,
the coefficient of friction between the
friction slab and the cohesionless soil
against which it has been cast, is less
than C (T=0).
0 In the absence of more
riy
specific information, y' may be taken
as being equal to tan 0 .
1
Until such
time as the behaviour of friction slabs
under earthquake conditions has been more
extensively investigated, the use of a
conservative strength reduction factor,
0 - 0.60 is recommended. On the basis
of the approximate analysis presented in
Fig C4.4, the effectiveness of friction
slabs in Seismic Zone A is in some doubt.
The use of a sliding friction
slab as an energy dissipating device in
either a cohesive or cohesionless soil
may be a feasible design proposition,
but until this concept has been proven
by test, friction slabs should be used
in a design to provide positive anchorage
only.
C4.5.2 The use of a coarse gravel blanket
is recommended in cohesive soils to provide
free drainage and as it is felt that the
friction characteristics of the cohesionless
material can be more reliably quantified.
C4.6.1 The soil parameters listed in
Table 4.2 are suggested typical values
for use in a preliminary seismic analysis
only. The implied equivalence between
the listed K, and E values for different
h s
types of cohesive soil is very approximate.
Note that the modulii of horizontal
subgrade reaction, values , listed in
Table 4.2 are constants independent of
pile diameter. It follows that the
coefficient of horizontal subgrade reaction
k^ = K^/D, the ratio between the horizontal
pressure between the pile and the soil and
the deflection produced by the pressure
application at that point, decreases
linearly with increasing pile diameter.
In his 1955 paper, Terzaghi •23 points 4
out that there is an erroneous conception,
1
widespread among engineers, that the
numerical value of the coefficient of
subgrade reaction depends exclusively on
the nature of the subgrade'. Terzaghi's
1
bulb of pressure concept, used to
1
illustrate the effect of pile width or
diameter on k, is shown in Fig C4.5(a).
In 1972 Broms 4.24 t e d that although
s t a
knowledge of scale effects on the value
of the coefficient of horizontal subgrade
modulus was not yet complete, Terzaghi s 1
assumption that k^ was inversely proport-
ional to D was normally used for design
purposes. B r o m s • also refers to test
4 24
results which show that the shape of the
pile cross section (in sands) has only a
very small effect on k^.
Recent research • ^ has indicated
4 2
that the assumption that k, varies inversely
to changes in D (in clays; is not strictly
true. In a theoretical prediction of a
full scale pile deflection test from self
boring pressure-meter data, Hughes et a l • *>, 4 2
while making use of the Terzaghi assumption
that k^ varies as the inverse of pile
diameter, emphasised that because of the
non-linear nature of the soil p/y
relationship, the dependence of k on the
level of soil strain must be taken into
account. In the evaluation of the soil
spring constants in a Winkler model, the
k, values used were the secant gradients
or the soil p/y curves appropriate to the
pile lateral diametric strains, y/D (as
outlined in Section 4.3.2.1(b)).
C4.6 .1.1 The normally assumed variation
in the modulus of horizontal subgrade
reaction with depth in uniform cohesionless
and cohesive soils is illustrated in Fig
C4.5(b).
C4.6.2 Methods by which soil p/y curves
may be obtained from laboratory tests of
samples are presented by Matlock • 4 4
(soft clay) and Reese, Cox and K o o p • 4 2 7
(sand). Methods by which the static
p/y curves can be modified to more closely
represent cyclic loading conditions are
also discussed in these papers. A
comprehensive summary of these and other
methods is given by Hughes, Goldsmith and
Fendall - .
4 19
With further development, the
determination of p/y curves from insitu
testing methods (i.e. Menard pressuremeter
or self boring pressurementer) is likely
to provide more realistic design parameters
than p/y curves determined from laboratory
test data.' An outline of the historical
background and recent advances that have
 Design seismic force
= C ^ x weight of superstructure
H
Lateral Pile lateral Coefficient of
diametric horizontal
Pile displacement
strain subgrade reaction
diameter Pressure, p

.
k = ±-
y

W= weight of friction (additional restraint from

D hi y,
slab + overlying fill passive soil pressure neglected)
Pressure , p
ny y
W xC (T=0) = Seismic force acting on friction nD ny, k = p
= '
kh

H
nD =
D h " ny, n
* slab and overlying fill
• S e i s m i c force from bridge assumed to be
superstructure In - phase
(worst c a s e )
W Xyu ' = Frictional resistance on underside of
friction slab T h e coefficient of horizontal subgrade reaction varies as the inverse
of pile diameter ( for a given level of pile lateral diametric strain).
- Applies to both cohesive and cohesionless soils.
Ideal restraint from friction slab under earthquake conditions
Terzaghfs " bulb of pressure" c o n c e p t ( Reference 4-23)
= W x / - Wx c
H y u ( T = 0 )

= w[y-c H / t <T=o)] (a ) Assumed Variation in the Coefficient of Horizontal Subgrade Reaction

with Pile Diameter

Dependable restraint provided by friction slab under earthquake

condition K
h =
k
h
D K
h• h k D

= jZf W // -C H / i ( T = 0)

Suggested values:
- strength reduction f a c t o r , 0 =0-70

- friction coefficient, = tan 0'

where 0 1
= effective soil internal angle of friction

Uniform cohesive soil Uniform cohesionless soil

( b ) Assumed Variation in the Modulus of Horizontal Subgrade Reaction

FIG.C4-4 LATERAL RESTRAINT PROVIDED BY A FRICTION
with Depth in Uniform Soil cn
S L A B UNDER E A R T H Q U A K E CONDITIONS CD
FIG. C 4 - 5
260
been made in pressuremeter development
are given in References 4.28 and 4.29.
It should be noted that in contrast to the
Menard pressuremeter, determination of
soil parameters from self boring pressure-
meter test results are less dependent upon
empirical correction.
C4.6.3 The suggestion that for cyclic
loading the soil stiffness parameters
should be reduced to 70% of their static
values is based on the recommendations of
ACI Committee 3 3 6 - . 4 5
C4.7.1 The following method of assessing
liquefaction potential has been shown to be
in extremely close agreement with data from
sites where liquefaction has occurred during
strong ground motion. The method is based
on Chinese research (as reported by S e e d - ) .
The critical value of the stand-
ard penetration test resistance N i c r i t
separating liquefiable from non-
liquefiable conditions to a depth of
approximately 15 m is determined by:
N . = N (1 + 0.125 (z - 3) - 0.05
crit
(z - 2) ) (C4.10)
w that where liquefaction has occurred it
where z is the depth to the sand layer
under consideration in metres, is
the depth to the water table below ground
surface, and N is a function of the shaking
intensity as follows:
Approximate MM intensity N(blows per 0.3m)
9 16
8 10
7 6
6 4
Where the N value for a particular sand layer
is greater than N . , liquefaction is
unlikely to occur unSer design seismic
C
conditions.
Equivalent Modified Mercalli
(MM) intensities for bridge structures
designed for given earthquake return
periods in the three Seismic Zones may be
estimated from the following:
Return I eriod
5
50 Year 100 Year 150 Year
Zone A, 1^ 8.0 8.6
Zone C, I c
5. 4 6.0
For Zone B, the equivalent MM
intensity in terms of the return period and
the geographic coefficient 6 (defined in
Fig 2.1, Section 2) may be obtained from:
I = 4.3 log 6
B 1 Q + I A (C4.ll)
For most sites where evidence
of liquefaction has been observed, the
critical layer in which liquefaction
is believed to have occurred has been
located at depths of less than 15 m
and the ground water level has been at
depths of less than 5 m.
C4.7.2 Remedial methods, by which
liquefaction potential of very localised
pockets of sand might be reduced, could
involve:
(a) Increasing in-situ soil density by
vibro-flotation or displacement
piling.
4 3 0
(b) Removal and replacement of materials
with liquefaction potential.
(c) Lowering the ground water level
by draining.
(d) Grouting.
Evidence from site observations
following earthquake attack has indicated
has generally involved the complete bridge
structure rather than isolated pier found-
ations . It has also been observed that
damage resulting from the liquefaction
of soil overlying bedrock is not necessarily
avoided by socketing the piles into the
bedrock.
C4.8 REFERENCES:
4.1 Ministry of Works and Development
"Highway Bridge Design Brief", MWD
Civil Division Publication, CDP
701/C, Wellington, New Zealand,
April 1975.
4.2 Taylor, P.W. and Williams, R.L.,
"Foundations for Capacity Designed
Structures", Bulletin of the New
Zealand National Society for Earth-
quake Engineering, Vol. 12, No. 2 ,
June 1979.
4.3 Davisson, M.T. and Prakash, S.,
"A Review of Soil/Pole Behaviour",
Highway Research Record No 39,
Highway Research Board, Washington
DC, USA, 1963.
4.4 Matlock, H., "Correlation for Design
of Laterally Loaded Piles in Soft
Clay", Offshore Technology Conference,
Paper No OTC 1204, 1970.
4.5 ACI Committee 336 , "Suggested
9.0 Design and Construction Procedures
for Pier Foundations", Journal of
the American Concrete Institute,
6.4 August, 1972.
4.6 Kubo, K., "Vibration Tests on a
Structure Supported by Pile Foundation",
Proceedings of the Fourth World
Conference on Earthquake Engineering,
Vol III, Chile, 1969.

4.7 Seed, H.B., Lysmer, J. and
Hwang, R. "Soil Structure
Interaction Analyses for
Seismic Response", Proceedings
of the ASCE, Journal of the
Geotechnical Engineering Division,
Vol 101, No GT5, May 1975.
4. 8 Berger, E and Pyke, R. "Simplified
Method for Evaluating Soil/Pile/
Structure Interaction Effects",
Proceedings of the Offshore
Technology Conference, Houston,
USA, Vol 2, Paper OTC 2954,
May 1977.
4.9 Davisson, M.T., "Lateral Load
Capacity of Piles", Highway
Research Record No 333 - Pile
Foundations, Highway Research
Board, USA, 1970.
4.10 Kocis, P., "Lateral Loads on
Piles", Bureau of Engineering,
Chicago, USA, 1968.
4.11 "ICES STRUDL II, the Structural
Design Language Engineering
User's Manual", Massachusetts
Institute of Technology.
4.12 Poulos, H.G. "Behaviour of
Laterally Loaded Piles; I -
Single Piles", Proceedings of
the ASCE, Journal of the Soil
Mechanics and Foundations
Division, No SM5, May 1971.
4.13 Blaney, G.W., Kausel, E. and
Roesset, J.M. "Dynamic Stiffness
of Piles", Proceeding of the
Second International Conference
on Numerical Methods in Geo-
mechanics, Blacksburg, USA,
ASCE, 1976.
4. 14 Novak, M. "Dynamic Stiffness
and Damping of Piles", Canadian
Geotechnical Journal, Vol II,
No. 4, November 1974.
4 . 15 Margason, E and Holloway, D.M. ,
"Pile Bending During Earthquakes
6th World Conference on Earth-
quake Engineering, New Delhi,
January 1977.
4 . 16 Patterson-Kane, K.J., "Computer
Analysis of Pile Groups, NZ
11
Engineering, Vol 33, No 4,
April 1978.
4 . 17 Prakash, S., "Behaviour of
Pile Groups Subjected to Lateral
Loads", PhD thesis, University
of Illinois, USA, 1962.
4 . 18 Davisson, M.T. and Salley, J.R.
"Model Study of Laterally
Loaded Piles", Proceedings of
the ASCE, Journal of the Soil
Mechanics and Foundation
Engineering Division, SM4,
September 1970.
4. 19 Hughes, J.M.O. Goldsmith, P.R.
and Fendall, H.D.W., "The
Behaviour of Piles Subjected
261
to Lateral Loads" University of
Auckland, School of Engineering
Report No. 178, September 1978.
4.20 Kuthy, R.A. et al, "Lateral Load
Capacity of Vertical Pile Groups",
Engineering Research and Development
Bureau Research Report 4 7, New York
State Department of Transportation,
April 1977.
4.21 Pidgeon, A.W. and Toan, D.V. ,
"Computer Program for the Elasto-
Plastic Analysis of Pile Groups",
Beca Carter Rollings and Ferner,
Auckland, March 1978.
4.22 Priestley, M.J.N. Evison, R.J. and
Carr, A.J., "Seismic Response of
Structures Free to Rock on their
Foundations", Bulletin of the New
Zealand National Society for
Earthquake Engineering, Vol.
11, No. 3, September 1978.
4.23 Terzaghi, K., "Evaluation of
Coefficient of Subgrade Reaction",
Geotechnique, Vol 5, No 4 , 1955.
4.24 Broms, B.B., "Stability of Flexible
Structures (Piles and Pile Groups)",
5th European Conference on Soil
Mechanics and Foundation Engineering,
Madrid, 1972.
4.25 Stevens , J.B. and Audibert, J.M.E.,
"Re-examination of p/y Curve
Formulation", 11th Annual Converence
of Offshore Technology, Houston,
USA, May 1979.
4.26 Hughes, J.M.O., Goldsmith, P.R.
and Fendall, H.D.W., "Full Scale
Laterally Loaded Pile Test at the
Westgate Freeway Site, Melbourne,
Australia: Load Deflection Predict-
ions and Field Results", University
of Auckland, Department of Civil
Engineering Report 190, November
1979.
4.27 Reese, L.C., Cox W.R., and Koop,
r
1
F.D. "Analysis of Laterally Loaded
Piles in Sand", Proceedings of the
Offshore Technology Conference,
Houston, USA, Vol 2, Paper No OTC 2080,
1974.
4.28 Baguelin , F , Jezequel, J.F . and
Shields, D.H., "The Pressuremeter
and Foundation Engineering", Trans
Tech Publications , Claustal,
Germany, 1978.
4.29 Hughes, J.M.O. , Wroth, C P . and
Windle, D., "Pressuremeter tests
in Sands", Geotechnique Vol 27 ,
No. 4, 1977.
4.30 Seed, H.B., "Soil Liquefaction and
Cyclic Mobility Evaluation for Level
Ground During Earthquakes", Proceedings
of the ASCE, Journal of the Geo-
technical Engineering Division,
Vol 105, No GT2, February, 1979.
 262
SEISMIC DESIGN OF BRIDGES
DETAILING FOR DUCTILITY AND DUCTILITY CAPABILITY
L.G. Cormack*, R. Park**, H.E. Chapman***
5.1 Detailing for Ductility
5.1.1 General
Flexural members, whether or not they
carry axial load, should be designed and
detailed to meet the minimum requirements
of the current New Zealand Codes of Practice.
5.1.2 Solid Concrete Members
Concrete structures should be designed
and detailed in accordance with DZ3101
1980 (reference - ) . This draft gives
s 1
provisions for the design and detailing
of solid reinforced concrete members that
may be subjected to yielding under the
action of the design earthquake.
5.1.3 Hollow Reinforced Concrete Section
Where hollow reinforced concrete
sections are designed to undergo ductile
yielding they shall be designed in
accordance with the above draft, DZ3101,
but special studies should be performed
to establish the appropriate dependable
member displacement ductility factor.
5.1.4 Structural Steel
Structural steel members should be
designed and detailed to conform to the
requirements of NZS 3404 * . 5 2
5.2 Ductility Capability
5.2.1 Reinforced Concrete
Solid reinforced concrete members
designed and detailed in accordance with
DZ310l • may be assumed to have an
5 1
available minimum member displacement
ductility factor of 8.
Where special studies are performed
and additional transverse reinforcing is
provided member displacement ductility
factors greater than 8 may be depended
upon.
5.2.2 Prestressed Concrete
Prestressed concrete members should
be the object of special study. For
design procedures refer to DZ3101 •
5.2.3 Structural Steel
Structural steel members complying
with NZS 3 4 0 4 - may, unless special
5 2
studies are performed, be assumed to have
a minimum member displacement ductility
factor of 8.
* Consulting Engineer, Auckland
**Professor of Civil Engineering University of
Canterbury, Christchurch
•••Ministry of Works and Development
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
SECTION 5
COMMENTARY - SECTION 5
C.5.1.3
At the present stage of development
it is recommended that studies of hollow,
thin walled piers be based on the detailing
and design methods developed for shear walls.
C 5.2 Ductility Capability
To establish the minimum seismic
horizontal shear force for which bridge
structures are to be designed requires
assessment of the overall structural
displacement ductility factor. This is
defined as the ratio of maximum displacement
under the design earthquake to the theoretical
yield displacement both measured at the
centre of mass.
The calculation of the displacement
ductility factor for a particular bridge
structure is discussed in Section 2. This
factor is derived from considerations of
the stiffness of the horizontal load
resisting structure and from the minimum
dependable ductility of those members that
form part of the load resisting structure
and which are required to undergo yielding
during the Design Earthquake.
If the bridge structure is designed
on the basis of no yielding under the action
of the design earthquake then a member
ductility factor of unity will be adopted.
In general however it will be advantageous
to design for extensive yielding to take
place under the design earthquake. The
maximum structure displacement ductility
factor that may be used in design unless
special studies are made is 6.
For solid reinforced concrete members
the amount of member ductility that can be
relied upon depends primarily upon the type
and disposition of reinforcing and the
average axial stress in the member.
The provisions of the Draft New
Zealand Standard Code of Practice for the
Design of Concrete Structures DZ 3101, are
the minimum provisions recommended for
members which may be subjected to yielding
under the design earthquake. Research
5
has indicated that these minimum provisions
should lead to dependable member displace-
ment ductility factors in excess of 8.
If special studies are performed,
however, and additional transverse rein-
forcing is provided, increased member
displacement ductility factors may be
justified.
Research is continuing into dependable
member ductility factors5.3,5.4^
 263

In particular work is continuing

on the NZMWD document CDP 810/A -
5 5

which, it is intended, will develop a

method by which the member displacement
ductility factor can be calculated for
particular cases involving reinforced
concrete sections.

The use of prestressed concrete or

structural steel members as the flexural
members that are required to undergo inelastic
deformations so as to resist the earthquake
reduced accelerations is not common in New
Zealand.

Provisions for their design are

however given in DZ 3101 and NZS 3404.
Even greater member ductility factors than
8 should be available for structural steel,
however care is essential in detailing so
that the possibility of premature buckling
or fracture is avoided.

C 5.3 References
5.1 Standards Association of New Zealand,
"Draft New Zealand Standard, Code of
Practice for the Design of Concrete
Structures" : 1980.

5. 2 Standards Association of New Zealand,

"Code for Design of Steel Structures,
(with commentary)" NZS 3404:1977.
5.3 Potangaroa, R.T., "Ductility of
Spirally Reinforced Concrete Columns
under Seismic Loading", University
of Canterbury, Department of Civil
Engineering, Report Number 79-8
February 1979.

5.4 Gill, W.D., "Ductility of Rectangular

Reinforced Concrete Columns with
Axial Load", University of Canterbury,
Department of Civil Engineering,
Report Number 79-1, February 1979.

5. 5 New Zealand Ministry of Works and

Development, "Ductility of Bridges
with Reinforced Concrete Piers",
Civil Division Publication CDP 810/4,
April 1975, 109pp.
264
SEISMIC DESIGN OF BRIDGES

SECTION 6

MECHANICAL ENERGY DISSIPATING DEVICES

R.W.G. Blakeley*, L.G. Cormack**, M.J. Stockwell***

CODE: C6.1 The system of "base isolation"

generally comprises two basic elements:
6.1 The following criteria are to be
satisfied for design of bridge (a) The structure is supported on
structures incorporating flexible flexible mountings to isolate
mountings and mechanical energy it from the greatest disturbing
dissipating devices. motions at the likely predominant
earthquake ground motion frequencies,
6.2 The performance of the devices and
used is to be substantiated by-
tests . (b) sufficient extra damping is intro-
duced into the system to reduce
6.3 Proper studies are to be made resonance effects and keep
towards the selection of suitable deflections within acceptable
design earthquakes for the limits.
structure, taking due account of
local site conditions. Flexible mountings include elast-
omeric and sliding or roller bearings.
6.4 The degree of protection against It should be noted that the properties of
yielding of the structural members the bearings have a significant influence
under the design earthquake is to on the response of the structure and
be at least as great as that the forces imposed on the substructure.
implied in these recommendations Information on the dynamic behaviour of
relating to the conventional elastomeric and sliding bearings is given
seismic design approach without e l s e w h e r e • S e v e r a l types of mechanical
c 6

energy dissipating devices. devices have been developed by the Physics

and Engineering Laboratory of the New
6.5 Where possible, the structure is to Zealand Department of Scientific and
be detailed to deform in a con- Industrial Research to provide the extra
trolled manner in the event of an damping required under (b) above through
earthquake greater than the hysteretic energy dissipation.
design earthquake.
Many bridges traditionally have
had one basic element of a base isolation
COMMENTARY SECTION 6: system, that is flexible mountings by means
of elastomeric bearings. There may be
C6.0 NOTATION: advantages in terms of reduced response
by incorporation of flexible mountings
k = post-elastic stiffness of in an otherwise monolithic structure,
db \.. , ,
dissipators plus elastic although this will only be beneficial where
stiffness of bearings the predominant earthquake ground motion
frequencies are in the short period range.
k , = elastic stiffness of The addition of mechanical energy dissipating
p
piers plus bearings devices to a bridge on flexible mountings
may have the advantage of reducing resonance
- force due to dissipator effects and keeping displacements within
at zero displacement acceptable limits.
ordinate
The following are bridge applications
W = weight of superstructure where incorporation of energy dissipating
devices in bridges is most likely to be
0 = capacity reduction factor effective:
(as per ref C6.4)
(a) in regions of high seismicity;

* Ministry of Works and Development, (b) mounted on a stiff substructure;

Wellington
** Consulting Engineer, Auckland (c) mounted on a substructure desired
*** Christchurch City Council to remain elastic.

The corollary is that energy dissipat-

ing devices are unlikely to be effective and
may even be a disadvantage in regions of
low seismicity or where mounted on a flexible
or flexurally yielding substructure. It
is therefore expected that the base
isolation system will be used most frequently
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR E A R T H Q U A K E E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980

for structures in seismic Zone A.
The main potential for economic
advantage lies in:
(i) Possible savings in abutment
separation requirements and
joint details as a result of
reduced superstructure deflections;
(ii) redistribution of seismic forces
on the substructure; for example
control of seismic forces through
energy dissipating devices at
strong abutments rather than by
ductile yielding of piers;
(iii) use of non-ductile forms or
components;
(iv) greater damage control.
C6.2 Detailed information on the design,
development and testing of mechanical
energy dissipating devices developed to
date is given in ref C6.1.
Design requirements for lead/
rubber devices specific to bridges include
allowance for lengthening and shortening
effects such as temperature variations.
Procedure adopted in the past Tias been:
cw
(a) The displacement of the superstruct-
ure at "design earthquake" loading,
and corresponding shear deformation
across the lead/rubber devices,
is estimated on the basis of
design charts;
(b) the thickness of bearing is chosen
so that the shear strain at "design
earthquake" loading is approximately
0.5;
(c) the size of bearing is selected
after design for allowable total
shear strains under combinations
of dead, live and overload, wind
and temperature ^.2^ with an
0
allowance for reduction in area
of the braring equal to the area
of the lead cylinder;
(d) the diameter of the lead plug is
estimated from the effective
yielding shear stress evident in
test r e s u l t s ' to give the
0 6 1
desired strength at the zero
displacement ordinate.
It should be noted that the
desirability of thick bearings, for
increased horizontal flexibility under
seismic loading, may conflict with the
need for sufficient vertical stiffness
to keep vertical vibration under live
load within the required limits for
liveliness. Some compromise between
these two objectives may be necessary.
It is anticipated that the addition of a
lead plug to a thick rubber bearing will
increase the vertical stiffness and reduce
any liveliness problems, but no test
information is available as yet to confirm
this.
265
Test evidence indicates that the
lead/rubber device will "creep" at load
rates corresponding to ambient temperature
variations and transmit considerably lower
forces, approximately 50% than those at
earthquake load rates.
C6.3 It is important that consideration
be given to the likely earthquake ground
motions at the site of the bridge. Where
conditions are such that predominant
frequencies of the ground motion are
likely to be in the long period range of
structures, for example where the structure
is sited on deep, flexible alluvium or
where the critical earthquake event may
occur at a considerable distance away
from the structure, a flexible mounting
system may detrimentally affect the
response of the structure ** • ^.
0
In such
circumstances the structure is likely to
be better off with energy dissipating
devices than without them because of the
extra damping, but as a design approach
a base isolation system should not be
adopted in this case.
Suitable design earthquakes may
be regarded as those which have response
spectra characteristics similar to the
elastic design spectra specified in Section
2 of these recommendations, except that
special consideration must be given if
local site conditions could promote long
period ground motions as discussed above.
C6.4 In suitable applications this
requirement may be achieved with significant
construction cost savings, particularly in
Zones A or B but unlikely in Zone C . That
is, the reduction in design forces on
members of the substructure more than
compensates for the extra cost of the
devices and associated details. The
extent to which the degree of protection
is increased above the minimum specified
in this section, if at a l l , to reduce the
anticipated frequency of earthquake induced
damage should be resolved with regard to
the client's wishes.
Assessment of forces on sub-
structure members may be made for common
types of bridge using available design
charts **•1.
0
For unusual or major
bridges, a dynamic time-history analysis
using realistic energy dissipator char-
acteristics will usually still be required.
The design charts were prepared, on the
basis of parameter studies, for structures
with and without energy dissipators where
the substructure is to remain elastic.
These charts are presented in ref C6.1
and cover the following cases:
(a) Elastomeric bearings only at
both abutment and pier;
(b) energy dissipators at abutment
only;
(c) energy dissipators at pier only;
(d) energy dissipators at both
abutment and pier.
Earthquake acceleration records
used are El Centro 1940 N - S , artificial
266
Bl and Parkfield. The charts may be used
to assess either longitudinal or transverse
response, or if desired response along an
axis inclined to the principal axes.
As an example, a bridge structure
with energy dissipators located only at
abutments and elastic restraint at the
piers is illustrated in fig. C 6 . 1 . Figs
C6.2 and C6.3 are design charts for this
case where the abutment is rigid, the
energy dissipator strength Q , = 0.05W,
and for the El Centro 1940 NFS and Bl
earthquakes respectively. The procedure
for use of each chart is as follows:
(i) calculate weight of superstructure,
W ibility for inspection and repair,
(ii) calculate combined stiffness of
dissipator plus elastomeric
bearings at abutment, k and
db f
determine k ^ / W /mm
(iii) calculate stiffness of pier plus
elastomeric bearings (or pier
alone where superstructure is
built-in to p i e r ) , k _, and
Kl
determine kp^/V? /mm
(iv) from top half of chart, determine
intersection of line and
kp^/W curve to give force on
abutment on vertical axis and
superstructure displacement on
horizontal axis
(v) determine force on pier by
either
(1) multiply superstructure
displacement derived from
(iv) above by the calculated
pier stiffness, k
or
•pb'
(2) from bottom half of chart,
determine intersection of
kp^/W line and k ^ / W curve.
It is proposed that charts similar
to figs C6.2 and C6.3 will be produced
based on the design spectra in Section 2.
C6.5 This requirement is regarded as
sound engineering practice in view of the
uncertainties in modelling and analysis
of the structure and in the characteristics
of ground shaking. However, it is
recognised that this will not always be
possible, particularly where non-ductile
structural forms or elements are used.
In general, where ductility can be sustained,
the anticipated lower ductility demand on
structures incorporating energy dissipating
devices means that simplified detailing
procedures appropriate for structures of
limited ductility would be satisfactory.
The required controlled post-yield
behaviour may generally be achieved by
provision of suitable margins of strength
between ductile and non-ductile members and
by attention to detailing, but without full
capacity design procedures. For example,
where forces in the sub-struetyre^are
calculated using design charts * or from
dynamic analysis, and where it is desired
that the structure remain elastic up to
"design earthquake" intensity, suitable
provisions are:
(a) substructure members capable
of ductile flexural yielding are
to be designed for a probable
flexural strength (based on a
capacity reduction factor, 0,
of 1.0 and yield strength of
reinforcing steel of say, 1.15
times the minimum specified) equal
to the calculated "design earth-
quake" moment;
(b) non-ductile substructure members ,
or members in which damage is
unacceptable because of inaccess-
or all members in shear, are to
be designed for a dependable
strength (based on appropriate
value of 0 • and minimum
C 6 4
specified material strengths)
equal to the force calculated in
that member at the "design earth-
quake" ;
(c) the separation details between
superstructure and abutment are
to allow for a deflection at
least of 1.5 times the values
calculated at the "design earth-
quake" ;
(d) special reinforcement requirements
in NZS 3 1 0 1 - for confinement
c 6 4
of concrete in bridge piers need
not be complied with. However,
good practice should be followed
in the detailing of the transverse
reinforcement to enhance ductility
in the potential plastic hinge
zones. The provisions for
design of shear and confinement
reinforcement for structures of
limited ductility in Chapter 14
of NZS 3101C6.4 or Section 7
of these recommendations, provide
a guide but may be conservative.
(e) Care should be taken in detailing
to ensure the integrity of the
structure during earthquake
shaking. Satisfactory seating
lengths or alternatively positive
horizontal linkages should be
provided between adjacent sections
of superstructure at supports and
hinges and between superstructures
and their supporting abutments.
C6.6 REFERENCES:
C6.1 Park, R. and Blakeley, R.W.G.,
"Seismic Design of Bridges",
Summary Volume 3, Road Research
Unit Structures Committee, New
Zealand National Roads Board,
Wellington, October 1978, 150 pp.
C6.2 " Highway Bridge Design Brief" ,
#
Ministry of Works and Development,
Civil Division Publication, CDP
701/D, Wellington, New Zealand,
September 1978.
C6.3 Priestley, M.J.N, and Stockwell,
M.J., "Seismic Design of South
 267
Brighton Overbridge - A Decision
against Mechanical Energy
Dissipators", Bulletin of the New
Zealand National Society for Earth- 0-35
quake Engineering, Vol 11, No 2,
June 1978, pp.110-120.

C6.4 -,"New Zealand Standard Code of

Practice for the Design of Concrete
Structures", NZS 3101, Standards
Association of New Zealand,
Wellington, to be published in
1980.

2 0-20

//AV/Ay// t 4
r

'
energy dissipator

(a ) Structure
////
// jj

l\\ 20 30 40 50 60 70 80
Superstructure Displacement, mm
90

F i

OdyT

^ /
j 7

abutment pier

tb) Force - Deflection Characteristics

0-20
FIG. C 6 . 1 : BRIDGE W I T H ENERGY DISSIPATORS
AT ABUTMENT ONLY.
FIG. C6.2: ENERGY DISSIPATORS ON RIGID A B U T M E N T ,
Q = 0.05W, E & C E N T R 0 1940 N - S .
d
268

SEISMIC DESIGN OF BRIDGES
SECTION 7
SMALL BRIDGES
R.W. Fisher*, A.G. Lanigan**, M.J. Stockwell***
7,0 NOTATION:
Basic horizontal seismic co-
efficient.
Q (T=0)= Basic horizontal seismic co-
efficient for period T = 0
seconds.
H = Seismic base shear force
T = Fundamental natural period
of structure
u = Structure ductility factor
7.1 INTRODUCTION:
Although procedures for assessing
design loading are given, the prime
consideration for design should involve
the selection of sound structural form
and careful detailing. At the lower
end of the small bridge range and for
small box culverts, it is envisaged that
attention to good detailing will by itself
provide all that is necessary for adequate
seismic resistance. As bridge size,
or design complexity increases, then the
need for some form of seismic analysis
assumes more importance. The designers
judgement in these matters must remain a
critical aspect of small bridge design.
7.2 OBJECTIVE AND DEFINITION:
The objective for small bridges
is to introduce seismic design concepts
which are suitable for application within
limited resource design offices and which
provide a similar standard of protection
to that required for larger structures.
For the purposes of seismic
design it is difficult to specify the
upper limit of small bridges or to place
them in definitive categories and
experienced engineering judgement will
be necessary in deciding the limits of
application for the design procedures
in this section.
7.3 DESIGN STANDARDS:
There is a greater total invest-
ment in the small size bridges group than
in the larger structures. When occurring
on important highway or rail links, earth-
* Assistant Bridge Engineer, Chief Civil
Engineers Office, NZR, Wellington
** Consulting Engineer, Babbage Partners,
Auckland
*** Christchurch City Council
BULLETIN OF THE NEW ZEALAND NATIONAL SOCIETY FOR EARTHQUAKE ENGINEERING, VOL 13 NO. 3 SEPTEMBER, 1980
269
quake damage to a number of small bridges
could well be collectively significant
in terms of public safety, disruption
to vital traffic and encomomics of
reinstatement.
For these reasons it is
desirable to adopt where possible a similar
design philosophy and standard of protection
to that used for the larger structures.
7.4 DESIGN CONCEPTS:
The nature of many small bridges
is such that their interaction with surround-
ing soils (such as at abutments) dominates
or significantly modifies their seismic
response compared with that of an equivalent
freely vibrating structure. As a simplified
concept many bridges of short length can
be visualised as being "locked into" the
approach embankments during seismic
activity. For this reason it may be
difficult to correctly model the structures
likely behaviour using conventional seismic
analysis.
Because of the expected soil
interaction dominance with all the associated
uncertainties, excessive emphasis on the
calculation of appropriate seismic loading
is not recommended. Instead, design
concepts for small bridges need to place
heavy reliance on sound detailing.
In general the design approach
described in 7,5 of this section
should be used for small bridges. However
where this approach precludes desired
refinements in structural proportions and
detailing, then a more comprehensive
analysis may be carried out in order to
achieve this.
An an alternative in certain
suitable cases, the use of energy dissipators
may be considered (see clause7.6)
7.5 DESIGN APPROACH ASSUMED
DUCTILITY VALUES WITH SPECIAL
EMPHASIS ON DETAILING:
7.5.1 General
For the purposes of "small"
bridges considered under this clause the
following should apply:
(1) The structure should be designed
to resist a seismic base shear,
using the procedures given in
Section 2.
(2) Each structure should be allocated
an assumed ductility factor
according to its primary mode of
seismic resistance as defined in7»5. 2.
 270
(3) Except where the primary mode of
resistance is as described under
(4) below, the basic seismic
coefficient should be obtained
by using the calculated structure
period. No allowance need be
made for foundation flexibility
unless seismic displacements
are required to be calculated.
(4) Where the mode of resistance is
provided by (a) friction slabs
(b) passive soil resistance at
the abutments or (c) tie back
systems at the bridge ends such
as "dead man" or soil anchors
then the basic seismic coefficient
should be obtained by taking
the structure period as zero i.e.
C (T=0).
u In the cases of
(a), (b) and (c) above, a safety
factor of 1.5 on assessed soil
strengths should be used in
conjunction with the seismic base
shear force (H).
(5) Where applicable, the earth
pressure component of earthquake
effect should be allowed for in
addition to the structure inertia
forces.
7.5.2 Ductility Values -
Following are the structure
ductility categories to be assumed for
design.
u — 6 Provision of energy dissipation
by moment resisting column type
pier shafts which will hinge
above ground or water level.
u = 4 Energy dissipation by foundation
hinging below ground, where hinge
zone will be reasonably accessible.
u = 3 Energy dissipation by:
(a) Foundation hinging below
water or ground level but
hinge zone in inaccessible
location.
(b) Non-yielding wall type piers
on footings.
u = 2 Piers with raked piles where
there is no provision for hinging
to occur in the pier.
7. 5.3 Detailing -
The "assumed ductility" design
approach requires attention to details
in order to avoid potential collapse
situations and to facilitate repairs after
earthquake damage.
The minimum requirements for
ductility detailing as defined in ref.7.5
should apply.
Foundations should receive special
consideration. In order to provide for
the inevitable situations where piles
hinge (whether by choice or accident) it
is advisable to select a pile type which
has the best possible deformation
characteristics. It is considered that
concrete piles detailed for ductility and
having substantial shear strength
throughout their length, would fulfill
the requirements of these situations.
Steel section piles would be deemed to
have the necessary ductility capabilities.
1.6 ENERGY DISSIPATORS:
In cases where is is economically
justifiable and in zones of high seismicity
the use of energy dissipating devices may
be considered for bridges having stiff
sub-structures. The requirements of
section 6 should be complied with.
7.7 OTHER CONSIDERATIONS:
7.7.1 Plastic Hinge Location
Notwithstanding the situations
described in"l.5. 3 every effort should be
made to locate potential plastic hinges
above ground or water level. In all
instances care is required in order to
preclude the occurrence of damage in
critical structural members. Capacity
design procedures are required.
7.7. 2 Integrity of Span to Pier
Connections -
Because many small bridges will
have simply supported spans the provision
of positive span to pier connections will
usually be an important design consider-
ation .
Span to pier connections should
have sufficient shear strength to satisfy
the "Capacity Design" principles as
described in Section 3. Where there is
difficulty in fully achieving this, then
a second line of defence such as providing
additional pier top width for span seating
should be considered.
7.7. 3 Design Ground Levels -
Due to ground level variations,
pier stiffness may be somewhat uncertain.
It is possible that in time, the ground
level may change around the piers of
certain bridges. This is particularly
likely in the case of river crossings.
In such instances, it is possible that
initially assumed design response char-
acteristics of the structure may alter
significantly. Conservative allowance
should be made for this contingency when
shear distribution among piers is being
evaluated.
7.8 CULVERTS AND VERY SMALL SINGLE
SPAN BRIDGES
As stated in7,4 the seismic be-
haviour of many types of smaller bridges
would be dominated by interaction with the
supporting soil. Culverts and small
single span bridges can be included in
this category.
Detailing to improve structural
resistance to ground deformation is of
importance. The arrangement of corner

reinforcement in culvert barrels, the
connection of all parts, such as wing
walls, headwalls etc., to the main
barrel and the provision of adequate
longitudinal reinforcement (or tying)
to resist differential settlements and
longitudinal separation, all require
attention in culvert design. For large
reinforced concrete box culverts and
underground pedestrian subways, the
design principles described in section 9
should be applied.
In the case of small culverts
and small single span bridges, it is
considered that seismic design should
consist almost entirely of sound detailing.
Corrugated metal culverts are
flexible conduits and as such are entirely
dependent upon their ability to deform in
accordance with the surrounding soil.
The only design recommendations which could
be made in these cases is that the use or
otherwise of this type of culvert should
be considered in the light of the likely
supporting soil deformations under
seismic action and the consequences of
possible failure.
COMMENTARY:
C7.5 DESIGN APPROACH : ASSUMED
DUCTILITY VALUES WITH SPECIAL
EMPHASIS ON DETAILING:
Ductility Values (referT.5.2)
The chosen ductility values
represent an attempt to categorise some
of the more common types of structure
in terms of seismic response and
serviceability. Background investigation
has indicated that the response of a
column hinging pier having ductility
reinforcement in accordance with DZ3101 may
generally be satisfactorily represented
by assuming a fixed base and u = 6.
Although the above assumption has been
demonstrated to apply only to piers
having periods of 0.3 sees. or greater
it is considered that in practice most
cases will not be affected by this
restriction, i.e. where structures have
shorter fixed base periods than 0.3 sees.
it is usually found that member sizes
are governed by practical rather than
seismic considerations.
In part, the chosen ductility
factors for the remaining structure
categories reflect a penalty for
undesirable or uncertain energy dissipating
mechanisms, rather than entirely indicating
the expected response; e.g. imposing a
higher design base shear on substructures
which hinge below ground or water level
will minimise seismic damage in inaccess-
ible hinge zones.
It is considered that in cases
where seismic resistance is being provided
by piers or abutments alone, (i.e. without
the assistance of passive soil resistance) ,
it would be unnecessary to design for a
ductility factor of less than 2 as most
structures should be capable of providing
this measure of ductility. Soil inter-
271
action effects such as plastic deformation
of foundation material and rocking will
contribute some measure of energy dissipat-
ion in the majority of the least ductile
structures.
Rocking Foundations (refer7.5 . 2)
References 7.1 and 7.2 indicate
that seismic protection can be achieved
with "rocking" footings. For small
bridges, it is considered that in the
absence of a "rocking pier" analysis,
non-yielding piers on footings may be
assumed to have an equivalent structure
ductility of 3.
A similar principle of protection
could perhaps be argued for pile foundations.
However, in this case, there are recognisable
difficulties with the prediction of force
levels necessary to induce the energy
dissipating "pumping action". This action
when occurring within controlled limits
in piled foundations would probably provide
some secondary benefits, but in these cases
it is not recommended that it be used as
the principal design means for dissipating
energy.
Raked Piles (referT.5. 2)
There have been reports of unfav-
ourable seismic behaviour with substructures
incorporating heavily raked piles. Due
to the largely axial resolution of
horizontal forces, this type of structure
will tend to have low ductility; hence
the recommended u = 2. However, where
provision is made for hinging in a pier
stem above the raked pile foundation, then
a reduced response would be achieved.
Also in cases where the rake is only
slight, a predominantly flexural type
response similar to vertical piles could
probably be obtained.
Propped Superstructure Between Abutments
(refer7.5.1)
Where there is negligible clear-
ance separating abutment walls from the
superstructure, bridge structures will
tend to interact with the approach
embankments during seismic activity. In
such cases provided slump failure of the
approach embankments is unlikely, then
passive soil restraint at the abutments
may be considered a satisfactory form of
longitudinal resistance for bridges of
short length.
Where relatively intimate contact
between superstructure, abutment and
approach embankment exist, then the
seismic energy input from both embankment
and foundations in combination with
passive restraint through an opposing
abutment, could cause complexities in
structural response. In the case of
rail bridges, the rails would introduce
further complications. Also there would
be some difficulty in accurately assessing
the strength and stiffness of the soil
behind the abutments.
Nevertheless, although the actual
seismic behaviour in these cases would be
272
difficult to assess theoretically, small
bridges of the type which inherently
interact with their approach embankments,
have quite often satisfactorily survived
previous earthquakes. This satisfactory
performance has occurred where approach
embankments have not slumped severely
(thus maintaining longitudinal bridge
stability) and where general detailing has
been sound. It is unlikely that this
type of structure would have in the past
been subjected to much more than nominal
only seismic design treatment.
For the purposes of introducing
a reasonable design method for obtaining
structure inertia forces it is to be
assumed that bridges which are "locked
into" the approach embankments will
respond seismically as "rigid" structures.
As such it may be further assumed that
where passive soil resistance is allowed
for at the abutments then the basic
seismic coefficient may be obtained by
using a period of zero seconds. To
prevent possible excessive movement which
may damage certain structural parts and
to allow for possible soil property
uncertainties a safety factor of 1.5
is specified on the assessed passive soil
strength.
The earth pressure component
of earthquake effects must be allowed
for in addition to the structure inertia
forces.
Where this mode of resistance is
to be relief upon, a wedge of selected
granular soil will need to be specified
for placing immediately behind abutments.
Friction Slabs (refer7.5.1)
Friction slabs could be used to
provide both longitudinal superstructure
resistance and lateral restraint for
superstructure diaphragm action.
However, many of the interaction
uncertainties as mentioned in the preceding
section on "Propped Superstructures" will
also apply to friction slabs.
If the force levels at which
friction slabs commenced sliding, could
be confidently predicted, then significant
energy dissipation could be allowed for
in design.
Because of uncertainties in this
regard, and to help allow for the attendant
risk to linkages and other connections,
design should be based on the assumption
that friction slabs will generate large
forces in the superstructure. Hence a
"rigid structure" response and safety
factor of 1.5 on assessed strength has
also been allocated to this mode of
resistance.
For similar reasons as apply to
propped superstructures and friction slabs,
tie back systems at bridge ends such as
"dead man" or soil anchors should also be
designed to "rigid structure" principles;
i.e. T = 0.
C7.7 OTHER CONSIDERATIONS:
Integrity of Span to Pier Connections
(refer7.7.2)
It has been shown that a significant
number of overseas bridge failures have
resulted from inadequate connections and
inadequate pier top seating. Ground
deformations can combine with inertia
effects in producing forces tending to
pull spans from piers. On bridges with
stiff substructures span to pier connections
could have to absorb large inertia forces,
It is a requirement of capacity
design to ensure that structural components
have adequate shear strength in order to
develop plastic hinges. A simple
extension of this principle would likewise
require that span to pier connections had
a similar shear capacity.
C7.8 CULVERTS AND VERY SMALL SINGLE
SPAN BRIDGES:
Culverts -
Damage to culverts due to lack of
structural integrity occurred in the
Inangahua earthquake * and it is considered
7 6
that the worst consequences of culvert
failure or the failure of culvert components
could be serious.
Inadequately designed head walls
or wing walls with poor connections to the
main barrel could allow part of the
embankment to collapse into and obstruct
the waterway. Stability of side slopes
on embankments in the vicinity of culverts
needs consideration. A flexible corrugated
metal culvert if subjected to extreme soil
deformation or loss of soil support could
collapse with a similar result. Corrugated
metal culverts may fail in a "snap through"
buckling fashion when cross sectional ^ ^
warping reaches 20% of their diameter.
C7.9 REFERENCES:
7.1 Priestley, M.J.N., Evison, R.J.,
Carr, A.J., "Seismic Response of
Structures Free to Rock on their
Foundations"; Bull. N.Z. National
Soc. Earthquake Engineering, Vol.
11, No. 3 September, 1978.
7.2 Taylor, P.W., Williams, R.L.,
"Foundations for Capacity
Designed Structures", Bull. N.Z.
National Society for Earthquake
Engineering, Vol. 12, No. 2
June, 1979.
7.3 McCullock & Bonilla, "Effects of
the Earthquake of March 27, 1964
on The Alaska Railroad,"
Geological Survey Professional
Paper 545-D.
7.4 Gray, P., Schofield, A.N.,
Shann, C D . , "Design and Construct-
ion of Buried Thin Wall Pipes",
C.I.R.A. Report 1978.
7.5 Standards Association (N.Z.)
"Code of Practice for Concrete
Design, DZ 3101".
 273
7.6 Duckworth, W.J.H., "Inangahua
Earthquake Damage on Railways,"
Bull. National Soc. Earthquake
Engineering, Vol. 1, No. 2
Dec. 1968.

Refer 7-5-1 KJO1"<L •• fS « Structure ductility

factor.

1 IP <n

0 U ILP W H-K hinges

Hinge, zona accessible. Whore hinae zone. is reasonably

accessible in piles

hincjas

MB5 M=2
VJhere hincje zona is inaccessible Rak.ed p\\cL interaction
below water or at considerable, Vtih soil
depth balow ground.

- Friction slab -Pass'^e soA resistance

Gm/d I T « O )
(assumed riajd restraint) (assumed ri^id restraint}

F I G . 7.1 A S S U M E D D U C T I L I T Y & S T R U C T U R E RESPONSE FOR S M A L L B R I D G E S

274
SEISMIC DESIGN OF BRIDGES
SECTION 8
STRUCTURAL AND NON STRUCTURAL DETAILS
A.G. Lanigan*, R.L. Preston**, R.W. Fisher***, M.J. Stockwell****
Ll GENERAL:
The following sections outline
various essential design considerations
for relative movement of structural
members, structural integrity and repair of
seismic damaqe.
8.2 DISPLACEMENT CONSIDERATIONS
8.2.1 Attention should be given during
design to the avoidance of damage to major
structural elements resulting from large
relative deformations induced by strong
earthquake motions.
Attention should also be given to
the design of clearances around minor
structural elements such as deck movement
joints. At such locations strong seismic
motion can be expected to cause damage.
The designer should therefore detail such
elements of the structure so that the damage
occurs in a predictable fashion with
permanent repairs being undertaken with
relative ease.
8.2.2 Clearances between major structural
elements and around items such as holding-
down bolts designed for relative movement,
may be calculated using Equation (2.3) or
(2.4) of Section 2.2.2 depending on
seismic zone. This equation can be used
to predict the maximum seismic displace-
ment of the centre of mass of a structural
system where dynamic characteristics
approximate those of a single degree-of-
freedom oscillator.
In cases where the structure cannot
be adequately modelled as a single degree-
of-freedom oscillator, it may be necessary
to resort to the use of more refined
analytical techniques in order to realist-
ically assess relative displacements under
seismic loading.
8.2.3 In structures of low displacement
ductility or in cases where the evaluation
of relative displacement is uncertain, it
may be necessary to use elastomeric
buffers to reduce possible impact forces
which may occur between major structural
elements during strong earthquake motions.
In addition, adequate clearances should
be left to ensure that large forces will
not develop during more frequently
expected moderate earthquakes.
* Consulting Engineer
** Ministry of Works and Development,
Wellington
*** N.Z. Railways
****Christchurch City Council
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
8.2.4 At points such as deck movement
joints, clearances in the joint and its
immediate supports after making allowances
for long term shrinkage and creep movements,
should be at least 0.15 and preferably
0.25 times the relative movement calculated
usina Equation (2.3) or (2.4) of Section
2.2.2.
8.3 STRUCTURAL INTEGRITY
8.3.1 Positive longitudinal linkage
should be provided between adjacent sections
of superstructure at supports and hinges
and between superstructures and their
pier supports.
At abutments positive horizontal
linkage between the superstructure and the
abutment should also be provided unless
the minimum overlap distances between
superstructure and substructure as defined
in figure 8.1 are satisfied.
Details of a variety of linkage
systems are given in figure 8.2. If the
linkage is at a point where relative
deflection between the sections of super-
structure or between superstructure and
substructure is intended to occur during
seismic motions, sufficient slack should
be left in the linkage so that it does not
start to act until the design seismic
deflection is exceeded.
8.3.2 Holding-down devices should be
provided at all supports or hinges in
continuous structures where the upwards
vertical reaction generated by a horizontal
or vertical seismic load opposes and exceeds
50% of the static dead load reaction.
In calculating the appropriate
upwards seismic design reaction at any
support or hinge, the horizontal seismic
force should be that required to form a
plastic mechanism, assuming all plastic
hinges have developed their overstrength
capacity. In all cases the minimum
design strength for the holding-down
device should be 20% of the dead load
downwards force which would be exerted
if the span was simply supported, or where
the net residual load is negative (dead
load minus upwards seismic reaction) use
twice the value of that load, whichever
is the greater.
8.4 REPAIR CONSIDERATIONS
8.4.1 The designer should consider the
likely method of repair and ease of access
to areas of a structure where seismic
damage will most probably be sustained.
 275

or 2 x A
(where A is obtained
from equation 2 - 2 - 2 )

FIG. 8 1 OVERLAP DISTANCE

FIG. 8 - 2 ALTERNATIVE ABUTMENT LINKAGE

276
8.4.2 The designer should consider the
hierarchy of inelastic failure of structural
members during strong earthquake motions.
It is recommended that members which are
expected to fail first during severe
seismic shaking should be the easiest
to repair both temporarily and permanently.
COMMENTARY SECTION 8:
C8.1 This section of work confines its
recommendations to detailing for displace-
ment rather than for strength requirements
which are covered elsewhere.
In order to speed up design methods
and^make structures more economical, the
MWD * have developed standard seismic
details for bridges which utilise their
simple supported standard pre-cast beams.
The detailing concepts incorporated in
these standard designs are nevertheless
applicable to the design of more complex
structures.
C8.2.2 The clearance requirements
described in this section are generally
only minimum values. The designer should
consider that clearances might also be
required (say between abutment backwalls
and the end diaphragms) for access for
maintenance and/or repair of seismic
hardware. MWD standards contain
recommendations on clearance for particular
structural layouts.
C8.2.4 Because actual movements are likely
to exceed those for which the deck joints
are designed it is accepted that damage
caused by strong earthquake motions may
occur in parts that are not main structural
members. As a result a plane of weakness
should be introduced to allow secondary
damage to occur in a predetermined and
limited manner, in order to permit early
use of the bridge after a major earthquake.
Judgement should be exercised in determining
the amount of long-term shortening which is
to be combined with earthquake movement
in deriving the design value for joint
displacement. Elastomeric bearings which
transfer significant seismic forces from
the superstructure to the substructure
should be positively anchored to their
supports with dowels or equivalent.
This system of anchorage is part-
icularly desirable where bearings could
possibly slide and fall off the bearing
seats under seismic action.
Elastomeric bearings should be
designed to meet the general requirements
of the DoE Technical Memorandum No. B E
l/76 8.l
c
t This document however, does not
specifically mention allowances to be made
in the design of a bearing under seismic
loading. Dynamic tests ^•2 ^
c
shown
a v e
that elastomeric bearings are able to
sustain several cycles of loading
reaching shear strains of 130% without
appearing to suffer catastrophic damage.
It is expected however, that
strains of such a magnitude would tend
to move the bearings on their seats (if
slipping is permitted) or in the case of
prolonged cyclic excursions to strains
in the order of 130%, bearings would
probably need to be replaced.
It is recommended that elastromeric
bearings should be designed to have maximum
permissible shear strains of 100% under
full design seismic load.
Expansion joints at supports,
which provide significant seismic restraints
to the superstructure should be designed to
fail in such a manner that minimum per-
manent damage is caused to the expansion
joint itself. MWD standard detailsC8.3
show expansion joints attached to cantilever
slabs which hinge up under severe earth-
quake loading. Alternatively at abutments
the expansion joint may be attached to a
deck slab which impinges upon the top of
the abutment backwall which in turn is
designed to be knocked off under severe
earthquake loading.
It is considered that the knock-up
cantilever slab shows most promise at this
stage.
The designer should consider the
use of expansion joints which can be supported
in such a way that the structural steel
components of the joint will receive little
or no damage severe seismic shaking, eg.
compression seals or strip seals.
C8.3.1 Structural integrity can only be
maintained if extreme displacements are
controlled to prevent any span elements
dropping from their supports.
Where possible, the superstructure
should be designed to be continuous or
linked together at pier supports with a
hinged linkage slab (normally as part of
the deck) to cause the deck to act as a
longitudinal diaphragm under transverse
seismic action but with no significant
influence on live load moment. This
diaphragm action can then be used to
maintain structural integrity and enable
the rational distribution of transverse
seismic forces between supporting piers
and/or abutments.
Figure 8.3 gives some details of
standard M W D * seismic linkage hardware,
C 8 3
eg. linkage bolts and their associated
buffer rings. Alternative reinforced
concrete shear key arrangements at abut-
ments and piers are also shown.
The provisions for calculation for
design loads in these linkage elements
are at this stage rather empirical in
nature. It is recommended however that
the design strength of the linkage element
should have a minimum value of 0.2 x the
weight of the heavier of the two adjoining
spans or parts of the structure, or such a
value as may be determined by a rational
analysis which takes account of the dynamic
interaction of the superstructure and
support elements.
C8.3.2 This section attempts to give the
designer some guidance about the design
of holding down bolts in continuous struct-
ures .
 277
ELXFKNlSiOM JOINT 4 0 0 r ^ M I N . FOIL ACCE\Sfs

LABOTV1KNT BO^KIQ - 4 0 0 ^ M I M . FOR. A C C E . »

& A \ C 1 C W A . L L WfTW KNOCK.
OFF o&vice.
c-LS
j i. A J b U T M C N T T BE-ARl
EXPANSION JOINT* UNCAGC PLATE.

S u r r i O M A l , ELE.W1QN THROUGH UIMK.AGL 5QLT

ALTERNATIVE ABUTMENT LINKAGE
( A L . T E . f t N / K T I V E . W I T H OZPK. CA\NITl L E - ' t R.
N

D C S I Q N 1 E . D - T O MINGS- U P W A S L O S . " )

SHEA.*, ice-v OPTIONAL-

FIG. 8 • 2
SLCTiONAL PLAN THROUGH LINKAGE. &OLTS
TYPICAL DETAILS OF SEISMIC
TYPICAL ABUTMENT LINKAGE
CONNECTIONS AT PIERS AND
ABUTMENTS

4 0 0 r ^ ^ M I M . FT>e ACCESS

SECTIONAL E.LEWION THROUGH LlNK>GL BOLT SE.CTONAL t.L£.VATiQM THKOUGH U N K - A G E A S S E M B L Y

TV8EJ2. SPACER'S // L£E.

W

HMCAsGE. PLATE.

L-.MK-AGE. P L - A C T ^
ejoe>^e. p a d s

•2.<SOr-^ MIN. U N L E S S _
2 G O r « m M I N - UNUE%S AUTELR.NAX.VE J A L T E e N A T i V E ACCESS SHEA*«„ K J E X O P T I O M A ^ L
0 ? T \ O SJ A t ,
access ISPEOViDCD. n* 1

5CCTSONAL PLAN "HH ROUGH LINKAGE BOLTS SECTORAL =>LAN '-ROUGH UNJUAGL A S S E M B L Y

TYPICAL PIER CONNECTION TYPICAL PIER CONNECTION

( With deck linkage ) ( With deck expansion j o i n t )

FIG. 8 • 2
278 f«y « • S V S E ^ f t AT P O S T V I E L O , A i ! 6 U M E f » y •> g » S WP®.
i r < ^ » sHEAa ^ t e e s s , assume ir#y » o - s s ^ fW^.
EWffiEOCBO Mfc - » E L A S T I C L I M I T frAOMEWT CAPACITY OP C H S ^W&AR. VCEV
MASS Me A B O U T AX«,X.-X A T
T Y P E L£*4STH
• S W E A R C A P A C T T V OP Q.W.S VCEV A TfcteWTAssJGLESTO A**SX-X

+ t . PLATB.» a e o x z o v BBS 0-9

£ LiNKLAG€ B O L T

804-8*304.-0* fe-O KW.m KJOAAiWAL R.HS S H E A R VCEY SEE M O T E S 3 & A.

ZOO fell - C G G E D MOLE.(p<=n-\OKiAL R D S
B O R E T U 6 E TO
c o m p l y wm-t b.s iB^n E . H S E H E A » KLEV T V P E 5 4- T O ©)
S H A L L B E PILL.EO WITH C O W C 2 . E T E
304-8*304 6 * l^-O i4»
- Twe d e s i g w o p

S H A L L T A K E /&OCOUMT
Oft
4 OFirwe p o r c e s
B Jpbom t h e e n s
- — S H E A R ycEY.
4 C e - 4 * "203-2 * ! * 2 j S
S T E E L CYLIWOEJ2L
crROM Awwr> PLATE
zoe FOE. &H<= SWEAR.
fcCEY TVPE&
S E E MCTE 3 .
13
A -
O U T u M E O P7

1-2. o-s COSED HOLE

•2O4 6 * 2 . 0 2 > - 2 . * 4© s
TMSGUGH PlEg CAP

B.H.S S M F A Q . WLEVS

LEKJCrw A 5 S.EjQ.U>S2jED

- W H E Q E K i E C E S S A e v WWCHiwE FLAT 2.QQ

TO DEPTH

\ - H O T C»P G A L V A N I S E D tSTEEL
BOLT BAR. X TM8.CAC T O V I E L D F C 3 C E
OlA MAO-UMEC FRO*A P L A j Kj B A R , T C M L S 3 4 0 2 R LOCKMUT
OE5GW*T\OU OTt R Kl
GQAOE t7S, NUT
zo KjOT 3EQO LI M I L A G E B O L T
a 52. 1 9 Q

AO E. AO 9-2. tA&c > a 1L<3 0

RUBBER RING BUFFERS


In determining whether or not hold
downs are required (refer first paragraph
of Section 8.3.2) the horizontal load
may be determined from Equation (2.1) of
Section 2.1.1. In looking at the reactions
caused by vertical seismic accelerations
the designer is referred to Sections 2.3,
C2.3.2 and C2.3.3 for methods of deter-
mining the magnitude of these reactions.
The recommendations contained in
Section 8.3.2 are rather empirical in
nature as they contain complex loadings,
ie. vertical reactions induced by horizontal
earthquake action, or alternatively, vertical
reactions from vertical earthquake action.
It is recommended however that the
designer need not consider the concurrent
effects of vertical and horizontal seismic
accelerations when designing holding-down
bolts.
The designer is urged to exercise
broad judgement in determining the size
and location of hoId-down bolts. In many
instances it is considered that the overall
capacity of the hoId-down bolts should be
at least twice that required by Section
8.3.2 where it may be extremely difficult
to replace such items in the event that
they fail under earthquake loading.
279
repair the abutment backwalls, diaphragms
or piles. In detailing the various
component members of this abutment, the
designer should therefore provide clear-
ances and member strengths which ensure
that the piles and backwall are the least
likely to sustain permanent damage.
C8.5 REFERENCES:
C8.1 Department of the Environment,
"Design Requirements for Elasto-
meric Bridge Bearings", DoE ,
Highways Directorate (Gt.
Britain), Technical Memorandum
(Bridges) No. B E 1/76, February
1976.
C8.2 Tyler, R.G. - "Dynamic Tests
on Laminated Rubber Bearings",
Bulletin of the NZ National
Society for Earthquake Engineering,
Volume 10, No 3, September 1977.
C8.3 Ministry of Works and Development,
"Standard Plans for Highway Bridge
Components", MWD, Civil Division
Publication, CDP/901, Wellington,
New Zealand, January 1978.

The designer should be aware that

in the unlikely event that a simply
supported span is subjected to an upwards
seismic acceleration only, then theoretic-
ally the hoId-down bolts would not be
required unless the upwards acceleration
exceeded gravity.
C8.4.1 This section refers for example,
to pier cap and abutment widths in cases
where the bearings might need replacement
after sustaining severe seismic shaking.
Sufficient clearance should be provided
to enable standard jacking equipment to be
used.

In regard to repair of plastic

hinge areas on piers for example, it may
be judicious to provide suitable locating
fixtures (or holes in the piers) from
which scaffolding could be suspended for
repair work.

C8.4.2 This section aims at making the

designer aware of the need to consider in
detail the mode of failure of the structure
in the event of strong seismic motion.

Consider for example the likely

seismic damage that could be incurred at
an abutment with the following character-
istics :

^Expansion joint at deck

^Linkage bolts through abutment backwall

and diaphragm between beams

*Elastomeric bearings
*Piles designed to take longitudinal and
transverse loading
It is suggested that it is easier
to repair properly detailed expansion joints,
linkage bolts and bearings, than it is to
 280

SEISMIC DESIGN OF BRIDGES

SECTION 9
EARTH RETAINING STRUCTURES
M.B. M a t t h e w s o n * , J . H . W o o d * * , J.B. B e r r i l l * * *

9.0 NOTATION earth force due to rotational

forcing of wall
a distance of roller from wall
earth force due to translational
b width of roller FU forcing of wall

B reaction at base of sliding wall inertia force on wall, abutment

or bridge superstructure
C, compressional wave speed in soil
maximum force transmitted from
peak ground acceleration coefficient superstructure to abutment via
force limiting connection
earthquake base shear coefficient
force due to at rest earth pressure
dead load 0
D
AP increment or decrement in at rest
displacement of centre of mass OE earth force due to earthquake
d
of wall
P
P force due to passive earth
Young's modulus for soil pressure

acceleration due to gravity AP decrement in passive earth force

PE due to earthquake
height of wall
"PE
P
P " A P
PE
at rest earth pressure coefficient
force due to static earth pressure
K* at rest pressure coefficient for
o unloading = P
S1' S2 P
two different values of P
Ko g

active earth pressure coefficient P weight of roller per unit width

active earth pressure coefficient R reaction at failure surface of

AE for combined gravity and earthquake sliding wedge
loads
T period of vibration
passive earth pressure coefficient
ZT sum of maximum tensions in
L live load reinforcing strips

n number of acceleration pulses displacement of forced wall

N threshold acceleration coefficient peak ground velocity

force due to active earth pressure a, b weight of abutment, bridge

P
A
AP increment in active earth force due s' w weight of soil, wall
AE to earthquake
y depth below top of wall
P = P + AP
AE A AE probability coefficient
Z
H
earth force due to compaction °<AE angle of inclination of soil
C
failure surface
AP. increment of decrement in static
earth force due to earthquake 'RE angle of inclination of failure
surface in reinforced earth
A P
E1' A P
E2 two different values of AP_
coefficient for location in
earth force against forced wall seismic zone B

Y unit weight of soil

* District Design Engineer, Ministry of
Works and Development, Christchurch Ay maximum response displacement
** Laboratory Services Engineer, M.W.D.,
Central Laboratories, Lower Hutt axial strain
*** Senior Lecturer in Civil Engineering, angle of rotation of wall
University of Canterbury, Christchurch
BULLETIN OF THE NEW ZEALAND NATIONAL SOCIETY FOR EARTHQUAKE ENGINEERING, VOL 13 NO. 3 SEPTEMBER, 1980

displacement ductility factor
Poisson s ratio for soil
1
numerical constant
earthquake induced curvature
e loads being exceeded, capacity design
orthogonal stresses at infinity
orizontal pressure
0 angle of internal friction of soil
angle of base friction
angle of wall friction
hi
9,1 DESIGN LOADS:
9.1.1 Pseudo-static methods of analysis
for determining earth pressure and struct-
ures displacements are recommended.
3*1.2 The peak horizontal ground accel-
erations and velocities to be used in
computing inertia forces and wall dis-
placements should be as follows:
TABLE 9.1
Seismic Zone
peak acceleration C g
Q H
peak velocity V (m/sec) 0.6Z
q H
Coefficients and 3 are given in table 2.1 and
figure 2.1.
9.1.3 The choice of design life and
earthquake return period, which determine
Z^, should be made in accordance with
section 1.3. Where a retaining wall
forms part of a bridge abutment and pre-
mature yielding or excessive displacement
of the wall would contribute to collapse
of the bridge, the same design life and
earthquake return period as used for the
bridge would apply. In most other cases
it is expected that the criteria set out
in section 1.3.3 for "Bridges of less than
average importance" would be applicable to
retaining walls.
9.1.4 Vertical earthquake accelerations
may be neglected.
9.1.5 The inertia force of the retaining
wall itself and inertia forces transmitted
to bridge abutments from the superstructure
should be included in the analysis.
9.1.6 Ideally all structural components
of a wall should be capable of resisting
the following load combinations without
exceeding the dependable strength of any
member.
U = 1. 3 5 D 4- 1.8 P s + 2.25 L
U = 1.35 D + 1.35 P s + (AP
U = 0.9 D + 1 . 3 5 P g + (AP + P
281
Transitory loads should be
omitted and/or any load factor taken
as unity if doing so produces a worse
effect.
To ensure satisfactory post-
elastic behaviour in the event of design
principles as set out in section 3
should be followed and potential plastic
hinge regions should be detailed to meet
the minimum requirements of section 5.
9.1.7 If permanent outward movement
during earthquake is not acceptable
the external stability of the wall
should be checked using load factor
equations (9.2) and (9.3) above in
combination with under capacity factors
for the soil not exceeding 0 . 9 .
If permanent displacement during
earthquake can be tolerated the load
factors in equations (9.2) and ( 9 . 3 )
may be taken as unity for the designed
mode of failure but a variation of at
least + 1 5 % on the probable soil
strengths should be allowed for to
determine upper and lower bounds on
threshold acceleration.
9.Z
A B C EARTH PRESSURES:
9,2.1 In general the total earth pressure
0.5Z .g 0.53Z g 0.15Z g during earthquake equals the sum of
Ri H
three components:
0.6BZ H 0.2Z H (a) Static pressure due to gravity
loads.
(b) Dynamic pressure due to earthquake.
(c) Forced wall pressure due to
displacement of the wall into
the backfill.
These soil pressure components
may be calculated by the following
methods:
(a) Elastic theory.
(b) Approximate plasticity theory,
e.g., Coulomb and Mononobe-Okabe.
(c) Computer analyses, modelling the
soil as Winkler springs or as
finite elements.
Figures 9 . 1 , 9.2 and 9.3 show
solutions for the components of earth
pressure against a smooth vertical wall
retaining cohesionless soil with a
horizontal ground surface and the water
table below the base of the wall.
A linear variation of soil pressure
with wall movement between the upper and
lower bound solutions may be assumed.
9,2,2 STATIC PRESSURES;
(9.1) 9.2.2.1 At Rest Pressure
E + P p + P )(9.2)
I At rest pressure, figure 9.1(a) ,
acts against a perfectly rigid wall.
E F + P ) (9.3)
x
 ff = 1-5 C * H
x 0

/
00
ro
P = I.K #H
0 0
2

where K = 1 - sin 0
Q
H / AP O E =

a = K *H
or K = 0
V
1- V
0-58 H
/
x 0
cr = 0-5C 2TH
x o

(a) AT REST PRESSURE (a) RIGID WALL PRESSURE

P = 1K *H'
A A

2
AP E
where K = 1 - sin 4 A A P , = 0-75 C t f r T 0

1 + sin 4
H
2
cr = K * H
x A

tr = 0-75C 2fH
x o

(b) ACTIVE PRESSURE

(b) STIFF WALL PRESSURE
a . b

cr = K * y
x 0 or K *y p Ko= 1
Ko \ a
PAE =
K = J_
p
COSM0-OC)
K A
\ where K = AE

\ COS C< + /sin 0 • sin ( 0 -

2

AP A E
V COS €<
hi \ and o< =
-1
3
\
°"x= ( K - K ) y H A E A

(c) COMPACTION PRESSURE (c) ACTIVE PRESSURE

FIG. 9-1 : STATIC EARTH PRESSURES FIG. 9-2 : DYNAMIC EARTH PRESSURES
 /
/
•failure plane

NW,
w
/
~~rr w /
II ' cr = K t f y
x P

li
II P9
F

II PE

// o- =0-8E e(1-^.)
x s

// 0b

(a) CANTILEVER WALL

(a) WALL ROTATED

failure plane

NW C T
Z Tmax.
J-
WsV
—r
RE
WW

cr =2-4Esij.
x

(b) WALL TRANSLATED

(b) REINFORCED EARTH WALL

FIG. 9-3 FORCED WALL PRESSURES

FIG. 9-4 : DISPLACEABLE WALLS
 284
9.2.2.2 Active Pressure
If the top of the wall moves out-
wards sufficiently to develop a fully
plastic state of stress in the soil the
pressure reduces to the active value
shown in figure 9.1(b).
9.2.2.3 Compaction Pressure
If backfill is compacted in layers
with a roller of weight pb travelling at
distance a from the wall the pressure
distribution shown in figure 9.1(c)
should be assumed. For a vibrating
roller pb equals the dead weight of the
roller plus the centrifugal force induced
by the roller vibrating mechanism.
9.2.3 DYNAMIC PRESSURES:
9.2.3.1 Rigid Wall
Figure 9.2(a) shows an approx-
imation to the earthquake induced pressure
on a perfectly rigid wall.
3-2,3.2 Deformable Wall
For a relatively stiff wall the
earthquake pressure shown in figure 9.2(b)
should be assumed. A movement of the
top of the wall of between 0.1%H and 0.2%H,
under the combined static and dynamic
thrusts, would be needed to obtain this
reduction from rigid wall pressure.
If the wall is sifficiently flexible
for the top to move outwards by at least
0.5%H, under the combined static and
dynamic thrusts, the active pressure
increment shown in figure 9.2(c) may be
assumed, but 0 . 3 3 H above the base should
be regarded as a lower bound for the
position of the resultant thrust.
9.2.3.3 Displaceable Wall
If a wall is free to move outwards
then subjected to a threshold acceleration
Ng less than the peak acceleration of the
design earthquake, C g , and the load -
Q virtual back of a cantilever or counterfort
displacement behaviour is essentially
rigid-plastic, the forward movement of
the centre of mass may be estimated
from:
3V N 9.3.3.1
d = (9.4)
C g
o^
If this permanent displacement is
acceptable the structure need only be
designed to withstand dynamic pressure
due to an acceleration Ng.
9.2.V FORCED WALL PRESSURES:
Figures 9.3(a) and (b) show the
earth pressures against walls subjected
to rotational and translational displace-
ments .
In the general case an adequate
estimate of the pressure against a forced
wall may be made by modelling the soil
as a system of Winkler springs.
It may be assumed that the pressure
at any depth does not exceed the passive
pressure but it should be noted that wall
friction may significantly increase the
passive pressure coefficient.
9.3 FREE STANDING WALLS:
3.3.1 Either of two approaches may be
taken to the design of free standing walls.
(a) Design the wall to remain elastic
and not suffer any permanent
displacement during the design
earthquake.
(b) Accept, but limit the magnitude
of, permanent outward movement of
the wall and design for a mode of
failure which avoids yielding of
structural elements wherever
practicable.
3.3,2 WALLS FOUNDED ON SOIL:
For gravity, counterfort and
cantilever walls founded on soil permiss-
ible permanent displacement should be
adopted as the prime criterion for
earthquake design.
A sliding, rather than a rotational,
mode of failure should be aimed for and
due account should be taken of the probable
variation in soil strength when estimating
the threshold acceleration to cause
movement and the wall displacement.
The relationship between threshold
acceleration and displacement given in
section9.233may be assumed.
All structural elements should be
designed in accordance with section9.1. 6 .
The forces to be considered in the
sliding block analysis of a cantilever
wall are indicated in figure 9.4(a).
Soil contained between the stem, base and
wall should be considered part of the wall
mass when calculating inertia forces.
9.3,3 WALLS FOUNDED ON ROCK OR PILES:
If yielding of structural elements
is to be avoided during the design earth-
quake, earth pressures and wall inertia
forces should be based on the peak ground
accelerations specified in sectional.2
and due account should be taken of the
stiffness of such structures in estimating
earth pressure.
9.3.3.2 To obtain significant permanent
outward movement it will, generally, be
necessary to allow yielding of the wall
stem or piles. The consequent loss of
serviceability, or the cost of removing
backfill and repairing the wall, may be
justifiable on economic grounds.
The minimum specified and probable
yield strengths of reinforcement should be
considered when estimating upper and lower
bounds for the threshold acceleration to

cause permanent deformation.
Section9,23-3may be assumed to
apply, but adjustment for wall rotation
is needed to determine the displacement
of the top of the wall as the formula gives
the displacement of the centre of mass.
9.3.4 TIED BACK WALLS :
9.3*4-1 If tie backs are restrained by
some form of deadman anchor and ties are
required to remain elastic during
earthquake, the peak ground acceleration,
from table 9.1.2, should be used to
calculate the earthquake earth pressure
due allowance being made for tie
flexibility.
The mode of failure, in the event
of overload, should be by yielding of the
ties rather than failure of the wall face
or connections between the ties and the
wall or anchor.
9.3.4-. 2 If tie backs are restrained by a
moveable anchor, such as a friction slab
designed to slide while the rest of the
structure remains elastic, limited
permanent displacement could be tolerated
with a consequent reduction in the
acceleration to be used in computing
dynamic earth pressure. The probable
variation in friction slab resistance
should be allowed for in determining
threshold accelerations and the strength
of ties and connections.
9.3.4-. 3 For walls of minor importance
permanent displacement resulting from
yielding of ties may be acceptable but
particular consideration should be given
to the post-earthquake effectiveness of
the tie corrosion protection system in
this case.
93,5 REINFORCED EARTH WALLS:
3.3.5.1 If the wall is to remain elastic
and not suffer any permanent displacement
the face panels and reinforcing strips
should be capable of resisting a dynamic
thrust in the range.
0. 75 C H 2
> AP^ *0.50 C T H ,
2
(9.5)
o .b o can be placed on the force transmitted
uniformly distributed over the height of
the wall.
9.3.5. 2 Alternatively the design may be
based on permissible permanent displacement
provided outward movement results from
pull out of the reinforcing strips or,
where this is impracticable, by ductile
extension of the strips.
If ductile extension of the tie,
rather than failure of the connection to
the face panel, cannot be guaranteed the
ultimate strength of the connection should
exceed the tie pull out force, based on
the probable apparent coefficient of
friction, by a factor of at least 2.
The formula given in section 9.2.3.3>
(9.4) may be used to estimate displacement.
285
To ensure that permanent
deformations are within the capacity of
facing panels to accept relative movement,
the displacement of the top of the wall
should not exceed 3% of the wall height.
The failure surface may be
assumed to be a plane passing through the
intersection of the wall face and the
ground surface in front of the wall
provided the top layers of reinforcing
are draped, or their pull out resistance
increased by other means, to prevent
premature displacement of the upper
section of the wall.
The angle of inclination of the
failure surface should be determined by
considering the equations of equilibrium
of the failure wedge with the ties acting
horizontally.
Upper and lower bounds on the
threshold acceleration required to produce
incipient failure should be calculated
by considering the ties acting (a)
horizontally and (b) along the failure
surface and allowing for probable
variations in the pull out resistance and
yield strength of the ties.
The forces to be considered in
the sliding block analysis of a reinforced
earth wall are shown in figure 9.4(b).
9.3.5.3 External stability should be
checked by considering active earthquake
earth pressure ( ) , acting against the p
A E
interior face of the semi-rigid reinforced
earth block. The peak ground acceler-
ation (C g) should be used for walls
o
designed to section33*5.1 and the upper
bound on threshold acceleration (Ng) should
be used for walls designed to section 9.3.5.2..
9.4 BRIDGE ABUTMENT WALLS:
9.4.1 FORCE LIMITING CONNECTION TO
SUPERSTRUCTURE
9*4*1.1 Where the connection is such that
the force transmitted between the super-
structure and the abutment is independent
of their relative movement an upper limit
(P ) and the abutment designed to with-
L
stand the forces shown in figure 9.5.
AP E and P^ should be assumed in phase.
P may, or may not, be in phase with
Li
AP„ and P . T
ill X
9.4.1.2 The magnitudes of P , AP and
O ill
P are dependent on the movement of the
abutment wall relative to the retained
soil. Limits on these forces are:
P p * P s + P F * P 0 + P (9.8)
c
 286

FIG. 9-5 LOAD LIMITING CONNECTION FIG. 9-6 : RIGID CONNECTION TO BRIDGE SUPERSTRUCTURE-
TO BRIDGE SUPERSTRUCTURE NO SIGNIFICANT INTERACTION WITH SOIL

AP Pi = C W I P
E1
0 b A P E

I
P

Psi
F

P S2

(a) DYNAMIC PRESSURES IN PHASE

AP E > AP E

Ps Ps

(b) DYNAMIC PRESSURES OUT OF PHASE

FIG. 9 - 7 : RIGID CONNECTION TO BRIDGE SUPERSTRUCTURE -

SIGNIFICANT INTERACTION WITH SOIL
 287

If the abutment moves towards (2) Dynamic components of earth

the soil it may be assumed that AP = AP pressure out of phase, figure 9.7
OE. (b) :
If the abutment moves away from
the soil P = 0. The peak ground It may be assumed that the struct-
acceleration C should be used to calculate ure remains stationary and is
Q

AP„ and P_ unless permanent outward subjected to rigid wall pressures.

movement of the abutment is permitted. AP, AP (9.9)

OE
9 4-2
r RIGID CONNECTION TO SUPERSTRUCTURE and P
S = p x P (9.10)
o +
*c
9, /f. 2 . 1 A rigid connection may be
hinged or moment resisting but no signific- 9.5 CULVERTS AND SUBWAYS:
ant relative horizontal movement can
develop between the abutment and the 9.5,1 AXIAL STRAIN
superstructure.
An upper bound on earthquake
The two basic categories of induced axial strain in a buried structure
bridges linked by rigid connections to is given by:
their abutments are -
± 1° (9.11)
(1) Systems in which the soil
pressures make no significant
change to the dynamic response
of the bridge. 9.5.2 CURVATURE
(2) Systems where the abutment soil An upper bound on earthquake
pressures have a strong influence induced horizontal or vertical curvature
on the dynamic response of the along the length of a buried structure
bridge. is given by:
9*4-2.2 No Significant Interaction o y
(9.12)
The maximum response displacement
Ay of the bridge superstructure may be
calculated as specified in section 2.2.2 9.S.3) STRESS ON CROSS-SECTION
ignoring the influence of the retained
soil, and earth pressures estimated 9.5.3.1 If the depth of soil over the
by applying this displacement to the structure exceeds the height of the
abutment. structure, earthquake induced stresses
on the cross-section may be determined by
The forces acting on the abutment applying the static orthogonal stresses
are shown in figure 9.6. Ay may, or at "infinity" as shown in figure 9.8(b)
may not, be in phase with AP . The limits and representing the soil between these
on P , AP„ and P_ are as stared in
0 planes of application and the structure,
b ill r as well as the structure itself, as a
system of finite elements.
section9A. 1.2.
The stresses at infinity are
9.4-.2.3 Significant Interaction defined by:
Two cases should be considered: +. V oEs (1 - S>
(9.13)
(1) Dynamic components of earth (1 + S> ) (l-2S»
pressure in phase,
The peak ground figure 9.7(a)
acceleration C :
q
V E
should be used to calculate the o s (9.14)
bridge inertia force and the (1 + V ) d-2-9)
dynamic components of earth
pressure. These stresses may be applied at
any orientation to the structure.
The period of vibration should be
calculated, taking the influence of the 9.5.3.2. If the soil cover is less than the
soil restraint on the overall system height of the structure the loads shown
stiffness into account, and the required in figure 9.8(a) should be used for the
displacement ductility factor for the design of rigid structures.
structure determined from figure 2.3, 2.3
or 2.4. 9.53.3 Flexible corrugated steel plate
structutes may be assumed to interact
The limits on P , AP„ and P^ n with the soil to produce a uniform
b Hi r distribution of earth pressure aroung the
are as stated in section 9.4-.1. 2 . For periphery.
initial calculation it may be assumed that:
P -f p 9.6 STRUCTURAL DETAILING:
SI A' "S2
r
0 C
Particular attention should be
AP El " A P
AE' AP E2 = A P
0E paid to the following points when detailing
288
reinforced concrete retaining walls:
(1) Laps in reinforcement at the
base of a cantilever wall
stem should be staggered.
(2) Joints subjected to opening
moments should have all main
bars anchored in a compression
zone and diagonal bars should be
provided across the joint to
control cracking.
(3) Piles should be reinforced for
ductile behaviour in the potential
hinging region immediately below
pile caps.
(4) Pile caps should be reinforced
to resist the loads imposed by
piles hinging.
COMMENTARY:
C9. 1 DESIGN LOADS
C9.1.2 C = C for period T
o Hy ^
0
each seismic zone.
C9.1.4 Vertical acceleration has a minor
influence on earth pressures and since any
peak horizontal acceleration may be
accompanied by an upward or downward
vertical acceleration there tends to be
a cancelling of effects as far as wall
displacement is concerned. •^2 A 9
nominal allowance for vertical acceler-
ation is made in the load factor equations
suggested in sectional.6.
C9.1.5 Permanent displacement of retain
ing walls has been observed following
earthquakes • / 9.8, 9.11 and the
9 7
influence of the wall inertia force on
displacement has been demonstrated
theoretically9.24 experimentally. 18
a n d
C9. 2 EARTH PRESSURES
C9.2.1 Only the simplest cases are
illustrated in figures 9.1, 9.2, 9.3 to
indicate the relative magnitudes of the
earth pressure components.
The effects of backfill slope,
wall friction, wall angle and water are
dealt with in reference 9.19.
C9.2.2 STATIC PRESSURES
C9.2.2.3 Compaction Pressure
The expressions for compaction
pressure given in figure 9.1(c) are
from reference 9.14. Ingold's solution
has the advantage of simplicity over the
methods proposed by Broms. -^ If the 9
wall can deform plastically, (eg. gabion
and cribwalls), as backfill is placed
compaction pressure may be insignificant.
C9.2.3 DYNAMIC PRESSURES
C9.2.3.1 Rigid and Deformable Walls
and
C9.2.3.2 The pressure distributions shown
in figures 9.2(a) and (b) are approximations
to solutions derived by Wood •^3 £ igi^ 9
and deformable walls assuming the soil
to be in an elastic state. Reference
9.33 contains pressure distribution
plots directly applicable to design of
rigid, rotating and cantilever walls.
Wood has shown that as the wall rotates
or deflects away from the retained
soil the magnitude of the dynamic
increment in pressure reduces and the
centre of pressure moves downwards towards
the lower third point of the wall. It
seems reasonable to assume a hydrostatic
distribution of pressure when the soil
reaches a fully plastic state but 0.33H
above the base should be regarded as the
lower limit for the position of the
active dynamic thrust as recent research
9.18, 9.32 suggests that the location of
the resultant thrust changes in a complex
way during ground shaking.
It would be prudent to assume
that A P A E acts at 0.4H above the base
when computing the overturning moment.
0 for The expressions for the active
dynamic thrust A P ^ given in figure 9.2(c)
E
are simplifications of the Mononobe-Okabe
formulae which are given in full in
references 9.19 and 9.29.
The movement of 0.5%H to achieve
an active state applies to cohesionless
soil only.
C9.2.3.3 Displaceable Wall
Outward movement may result from
sliding, tilting or yielding of the wall
structure.
The sliding displacement mechanism
is described by Richards and Elms *24 9
and the general validity of the theory
9 has been verified experimentally by Lai
and Berrill. - Studies of the
9 18
tilting displacement problem are
continuing.9.1,9.32
If a sliding block is subjected
to a single triangular acceleration pulse
with a peak of C g and the block is
Q
displaced when the acceleration exceeds
Ng the relative movement of the block
is given by:
V N
c~ 2 - 1_
2C a 12
o-
for C < 0.59 (C9.1)
where V" is the peak ground velocity,
O
(C9.2)
2C g
o^
As shown in figure C9.1, a
satisfactory fit to the upper bound
displacement of a sliding block subjected
to recorded natural earthquake ground
o r r

motions is obtained when n = 6.
The sliding block model is a gross
simplification of real wall- soil systems
but the expression given in section 9.2.3>„3
should predict the correct order of
magnitude of wall displacement and further
refinement is not justified at this stage.
C9.2.4 FORCED WALL PRESSURES
The approximations given in
figures 9.3(a) and (b) are based on
solutions derived by Wood °33 j 9
Tajimi 31 for soil in an elastic state.
9e
Using Winkler springs to simulate
the interaction between wall and backfill
Scott • 9
obtained similar results.
28
Ghahramani and Clemence •
give solutions for the dynamic passive
pressure for wall translation and rotation
about the top and bottom of the wall.
C9.3 FREE STANDING WALLS
C9.3.1 The first method will ensure a
more serviceable structure but for some
types of wall it may not be feasible or
may be uneconomic.
The second method is a more realistic
and economical approach but if outward
movement cannot be obtained without yielding
of structural members the cost of repairs
to restore serviceability may be unaccept-
able .
C9.3.2 WALLS FOUNDED ON SOIL
The behaviour of gravity walls
is discussed in references 9.6, 9.18, 9.24,
9. 32.
The Manonobe-Okabe solution for
P , figure 9.4(a), is given in reference
9.29.
C9.3.5 REINFORCED EARTH WALLS
C9.3.5.1 References 9.4, 9.20 and 9.23
with analysis and design of reinforced
earth structures for static and earthquake
loads.
Static earth pressures based on
measurements of stress within reinforced
earth structures are generally used for
design * but it is suggested that
specific consideration be given to
compaction pressure, as indicated in
sectionS.XZ.-^as it is the upper sections
of the wall which are most vulnerable
to damage during earthquake.9.2,9.26
The design earthquake loading
generally specified - is an adaption
9 20
of an empirical approach proposed by
Richardson • on the basis of a dynamic
9 25
test of a full scale w a l l . • ^ Implicit 9
in Richardson s method is acceptance
1 the load limiting connection would have
of significant permanent outward movement
of the wall. If permanent movement is
to be prevented dynamic pressure similar
to that on a relatively stiff conventional
289
wall should be allowed for.
A uniform distribution of
pressure is recommended as this approximates
the distribution observed in the full
scale t e s t • and is consistent with
9 26
the distribution of pressure against
a relatively stiff conventional wall.
C9.3.5.2 A rational method for estimating
the threshold acceleration for outward
movement is described by Bracegirdle. * 9 2
The limit of 3%H on permanent displacement
a n ( is based on the report by Richardson
et a l -
9
that their test wall tilted
2 6
about 5% but there was no other observable
damage or ill effect.
C9.4 BRIDGE ABUTMENT WALLS
Damage to bridge abutments during
y 1Z
earthquakes has been well documented. •^ 9
9.8,9.11
Stability of the approach embank-
ments should be carefully considered • ^ 9 2
as slumping of batters, settlement or
bodily movement may have a profound
influence on abutment performance.
The interaction of the bridge -
abutment - soil system is complex • 9
but rational bounds can be put on earth
pressure and displacements if the extent
of interaction between abutment and
superstructure is considered. •6 9
C9.4.1 FORCE LIMITING CONNECTION TO
SUPERSTRUCTURE
C9.4.1.1 A P.T.F.E. bearing or an energy
abosorber with rigid-plastic character-
istics are examples of load limiting
connections.
An elastomeric bearing is not
strictly a load limiting connection but
the maximum force which can be trans-
mitted through this type of bearing is
generally known and this upper bound
force can be used in the analysis.
Generally movement of the
superstructure away from the soil will
impose critical loads on the abutment
foundations and movement towards the
soil will impose critical soil pressures
on the abutment walls.
Most abutments that are not
rigidly connected to the bridge super-
structure will be permanently displaced
towards the superstructure during
earthquake. This movement should be
added to the estimated maximum response
displacement of the superstructure to
determine the required seismic gap.
P T and AP_ are not necessarily
Li -Ci
2 in phase because the force transmitted via
a time-history dominated by the natural
frequency of the bridge whereas the
dynamic soil pressure will tend to follow
the ground acceleration in magnitude
and direction.
290

C9.4.1.2 If the abutment is forced towards unlikely, and

the soil and the soil remains essentially-
elastic , (2) slippage between the structure
and the soil is likely to occur.
E = 0E
A P A P (C9.3)
Development of frictional
If movement is sufficient to resistance and slippage is discussed in
develop full passive pressure and the soil reference 9.30.
is being accelerated away from the wall,
Values for V are given in
P + P AP. P
- AP (C9.4) section.9.1.2 and typical values for C,
S F E P PE PE
The Mononobe-Okabe solution for
9. 29 gives A P approximately the same Compact granular soil 500 m/sec
"PE
p E

magnitude as A P ^ for average values of silty sand 300 m/sec

C and 0. Dili

o medium clay 150 m/sec

C9.4.2 RIGID CONNECTION TO SUPERSTRUCTURE
C9.4.2.2 No Significant Interaction
A shallow, spill-through type,
abutment propping a heavy superstructure
with piers providing lateral resistance
may fall into this category.
C9.4.2.3 Significant Interaction
Portal type bridges are typical
of this category.
The loading shown in figure 9.7(a)
is appropriate for a bridge founded on rock,
since movement of the surrounding soil
relative to the rock will occur during
earthquake, but it is conservative for a
bridge founded in soft soil, since in
this case the whole structure will tend
to move with the soil.
Typical portal bridges are likely
to have relatively short periods of
vibration initially and the soil-structure
system may stiffen further as soil is
consolidated against the abutments during
shaking. Hence response may be close
to the peak of the spectrum so that
adequate ductility is required for survival.
Expression (9.11) is also
applicable to above ground structures and
may be used to estimate the relative
displacement of bridge piers during
earthquake.
C9.5.2 CURVATURE
Expression (9.12) gives the
curvature arising from a shear wave travell-
ing in a direction along the length of
the structure.
This is also an upper bound
solution because it is based on the
assumption of no slippage between the
soil and the structure.
C9.5.3 STRESS ON CROSS-SECTION
The method of applying static
stresses at "infinity" appears to be
accepted in Russia.9.9 ,9.15 Duns and
Butterfield 9.5 have shown that static
solutions can give good approximations to
dynamic wave propogation problems.
Alternative forms for expressions
(9.13) and (9.14).

C9.5 CULVERTS AND SUBWAYS 61 = +— g^ c, V

d o (C9.5)

The expressions given in this ^ c„ V (C9.6)

section have been derived from elastic
wave propogation theory. The application
of elastic wave theory to estimate
earthquake strain and stresses on under-
ground structures is covered in references
9.5, 9.9, 9.13, 9.15, 9.17 and 9.30.
Abrupt differential ground movements
that could be produced by faulting are
likely to produce more severe effects.
Methods of analysing long underground
structures for fault movements are given
in reference 9.16.
C9.5. 1 AXIAL STRAIN
Expression (9.11) gives the strain
arising from a compressional wave travelling
in the direction of the longitudinal axis
of the structure. This is an upper
bound solution because:
(1) Superposition of peak particle
velocities and ground accelerations
over a long length of structure is
g d o (1 - v)
The expressions given for <$u, ^5v~
are upper bound solutions for the reasons
given in section C9.5.1.
Where the finite element method
is used it is recommended that the
"infinity" stresses be applied at a
distance of at least twice the effective
"diameter" from the structure, as shown
in figure 9.8. Where this is not
possible, because of the proximity of the
ground surface, the stress load falling
above this boundary should be applied
at the surface.
Solutions for the stresses in
circular cylinders are given in
references 9.5 and 9.27.
C9.6 STRUCTURAL DETAILING
Corner joints subjected to tension
on the inside faces occur between wall
WAWA
Wc

A P OE APOE

Po Pc +
Po Pc
+

(a) SHALLOW OVERBURDEN

(b) DEEP' OVERBURDEN

FIG. 9-8 : UNDERGROUND STRUCTURES

292
bases and stems and between abutments
and wingwalls. Adequate detailing is
necessary to ensure that the ultimate
strength of the members can be developed
and cracks on the earth face are well
distributed. Recommendations for
detailing based on laboratory tests are
given in reference 9.22.
C9.7 REFERENCES:
9.1 Berrill, J.B. and Elms, D.G.
"Seismic Response of Retaining
Walls and a Brief Survey of Soil
Structure Interaction", Proc,
N.R.B. Seminar on Bridge Design
and Research, Auckland, New
Zealand, Nov. 1978.
9.2 Bracegirdle, A. "Seismic
Stability of Reinforced Earth
Retaining Walls". To be
published in Bulletin of New
Zealand National Society for
Earthquake Engineering.
9.3 Broms, B. "Lateral Earth Pressures
due to Compaction of Cohesionless
Soils", Proceedings, Fourth
International Conference on
Soil Mechanics, Budapest, Hungary
1971.
9.4 Cormack, L.G., Blakeley, J.P.
Toan, D.V. and Goldsmith, P.R.,
"Reinforced Earth Retaining
Walls with Specific Reference
to Earthquake Resistant Design",
Project 4787, Structures Committee
R.R.U., National Roads Board,
New Zealand, April 1980.
9.5 Duns, C.S. and Butterfield, R.
"Flexible Buried Cylinders :
Part I - Static Response : Part II
Dynamic Response", Int. Journal
of Rock Mechanics and Mineral
Science", Vol. 8, 1971.
9.6 Elms, D.G. and Martin, G.R.
"Factors Involved in the Seismic
Design of Bridge Abutments",
Workshop in Research Needs Related
to Bridges, Applied Technology
Council, Palo Alto, California,
Jan. 1979.
9.7 Ellison, B.K. "Earthquake Damage
to Roads and Bridges - Madang,
R.P.N.G. - Nov. 1970", Bulletin
of the New Zealand National
Society for Earthquake Engineering,
Vol. 4, No. 1, 1971.
9.8 Evans, G.L. "The Behaviour of
Bridges Under Earthquakes"
Proc. New Zealand Roading
Symposium, Victoria University
of Wellington, N . Z . , Vol. 2, 1971.
9.9 Fortieva, N.N. "Analysis of Non-
Circular Tunnel Linings Located
in Seismic Regions", Soil
Mechanics and Foundation Engineer-
ing, Vol. 13, No. 3, May/June
1976.
9.10 Franklin, A.G. and Chang, F.K.
"Permanent Displacements of
Earth Embankments by Newmark
Sliding Block Analysis",
Misc. Paper S-71-17, U.S.
Army Engineers Waterways
Experiment Station, Vicksburg,
Miss., 1977.
9.11 Fung, G. et al "Field Invest-
igation of Bridge Damage in the
San Fernando Earthquake",
Report of Bridge Dept., Division
of Highways, California Dept.
of Transportation, Sacramento,
California, 1971.
9.12 Ghahramani, A. and Clemence,
S.P. "Zero Extension Line
Theory of Dynamic Passive Pressure
Journal of the Geotechnical
Engineering Division, Proc.
A.S.C.E., Vol. 106, No. GT 6,
June 1980.
9.13 Hall, W.J. and Newmark, N.M.
"Seismic Design Criteria for
Pipelines and Facilities".
Journal of the Technical
Councils of A.S.C.E., T.C.I.,
Nov. 1978.
9.14 Ingold, T.S. "The Effects of
Compaction on Retaining Walls",
Geotechnique Vol. 29, No. 3, 19 79.
9.15 Karzevadze "Earthquake Design
for Road and Other Structures",
1974. (In Russian).
9.16 Kennedy, R.P., Chow, A.W and
Williamson, R.A. "Fault Movement
Effects on Buried Oil Pipeline",
Transportation Engineering Journal,
A.S.C.E. , T.E. 5. ~, Sept. 1977.
9.17 Kuesel, T.R. "Earthquake Design
Criteria for Subways". Journal
of the Structural Division,
A.S.C.E., S.T.6., June 1969.
9.18 Lai, Cho Sim and Berrill, J.B.
"Shaking Table Tests on a Model
Retaining Wall", Bulletin of the
New Zealand National Society for
Earthquake Engineering, Vol. 12,
No. 2, June, 1979.
9.19 Ministry of Works and Development
New Zealand "Retaining Wall
Design Notes", C D . P . 702 C, July
1973.
9.20 Ministry of Works and Development
New Zealand "Reinforced Earth -
Notes on Design and Construction"
Supplement to C D . P . 702 C, 1973,
June 1980.
9.21 Newmark, N.M. "Effects of
Earthquakes on Dams and Embankments",
Geotechnique, Vol. 15, No. 2
June 1965.
9.22 Nilsson, I.H.E. "Reinforced
Concrete Corners and Joints
Subjected to Bending Moment"
National Swedish Building Research,
 293
Stockholm, 1969. 9.34 Zarrabi-Kashani , K., "Sliding
of Gravity Retaining Wall
9.23 Prendergast, B.B. and Ramsay, During Earthquake Considering
G. "Design of Reinforced Earth Vertical Acceleration and
for New Zealand Conditions", Changing Inclination of Failure
Third A.N.Z. Geomechanics Surface", S.M. Thesis, Dept.
Conference, Wellington N.Z. May of Civil Eng., M.I.T., Cambridge,
1980. Mass., 1979.
9.24 Richards, R. and Elms D.G.
"Seismic Behaviour of Gravity
Retaining Walls", Journal of
the Geotechnical Engineering
Division, Proc. A.S.C.E., Vol.
105, No. GT4, April 1979.
9. 25 Richardson, G.N. "Earthquake
Resistant Reinforced Earth
Walls", Proc. Symposium on
Earth Reinforcement, Pittsburgh,
April 1978.

9. 26 Richardson, G.N. Feger, D.,

Fong, A. and Lee, K.L., "Seismic
Testing of Reinforced Earth
Walls", Journal of the Geotech-
nical Engineering Division,
Proc. A.S.C.E., Vol. 103, No.
GTl, Jan. 1977.
9.27 Savin, G.N. "Stress Concentration
Around Holes", Pergamon Press ,
London, 1961.

9.28 Scott, R.F. "Earthquake - Induced

Earth Pressures on Retaining
Walls", Fifth World Conference
on Earthquake Engineering, Rome,
Vol. 2, 1973.

9.29 Seed, H.B. and Whitman, R.V.

"Design of Earth Retaining
Structures for Dynamic Loads",
Proc. A.S.C.E., Specialty
Conference, Lateral Stresses in
the Ground and the Design of
Earth Retaining Structures,
Cornell University, N.Y., June
1970.

9. 30 Shah, H.H. and Chu, S.L. "Seismic

Analysis of Underground Structural
Elements", Journal of the Power
Division, A.S.C.E., P.0.1,
July 1974.

9 . 31 Taj imi, H. "Dynamic Earth Pressure

on Basement Walls in the Elastic
Ground", Proceedings of the Third
Japan Earthquake Engineering
Symposium, Tokyo, Nov. 1970.

9. 32 Whitman, R.V. "Dynamic Behaviour

of Soils and its Application to
Civil Engineering Projects". Pan
American Conference on Soil
Mechanics and Foundation Engineering
Lima, Peru, Dec. 1979.

9 . 33 Wood, J.H. "Earthquake-Induced

Soil Pressures on Structures",
Report No. EERL 73-05, Earthquake
Engineering Research Laboratory,
California Institute of Technology,
Pasadena, California, Aug. 1973.
294
J.H. Wood*, R.W.G. Blakeley* M.J.N. Priestley**
10.0 NOTATION
= damping matrix
elastic and post yield stiffnesses
k
l' 2
k
respectively of the idealized
moment-curvature relationships
K = stiffness matrix
M = bending moment
M = moment at first cracking of
concrete
= moment depth of non degrading
hysteresis loop for prestressed
concrete
= theoretical ultimate moment
based on ideal strength
M = mass matrix
a = Rayleigh damping scalar factor
for mass
= Rayleigh damping scalar factor
for stiffness
= equivalent viscous damping ratio
= overall structural
ductility factor
= section curvature
= curvature at first cracking of
concrete
= section curvature when tension
reinforcement reaches full yield
* Ministry of Works and Development,
Wellington.
**Reader in Civil Engineering, University
of Canterbury, Christchurch.
B U L L E T I N OF T H E NEW Z E A L A N D N A T I O N A L SOCIETY FOR E A R T H Q U A K E E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
SEISMIC DESIGN OF BRIDGES
SECTION 10
DYNAMIC ANALYSIS
r
10.1 BRIDGES REQUIRING DYNAMIC ANALYSIS
Dynamic analysis to obtain maximum
earthquake forces and displacements or
ductility demand should be carried out for
bridges that fall in one or more of the
following categories;
(a) Major bridge structures considered
to be of particular importance.
(b) Mass of any pier, including any
allowance for hydrodynamic effects,
greater than 20% of the mass of the
superstructure estimated to contribute
to the inertia loading on the pier.
(c) Bridges with superstructure joints
that are designed to allow significant
relative movement between the
separated spans.
(d) Structures in which the lateral load
resistance is provided by structural
systems other than conventional piers
and abutment structures.
(e) Suspension, cable stayed or arch
bridges.
10.2 METHODS OF ANALYSIS
10.2.1 Elastic Behaviour
displacement
If the lateral load resisting elements
remain essentially elastic under the design
earthquake loads then the elastic modal
spectral analysis method should be used.
Modal responses should be computed using
the design elastic response spectrum
given in Section 2, and the total maximum
responses should be computed using the
square root of the sum of the squares
(SRSS) method.
10.2.2 Moderate Inelastic Behaviour
If the overall structural displacement
ductility factor y is less than 2 under the
design earthquake, the elastic response
spectrum method, as detailed in 10.2.1,
should be used by adopting equivalent overall
stiffnesses and viscous damping values.
10.2.3 Inelastic Behaviour
Where the overall structural
displacement ductility factor exceeds 2,
the inelastic time~history method in which
the response is computed using numerical
integration, should be used.
10.3 LOADING DIRECTIONS
Dynamic analyses should be undertaken
for the two principal horizontal directions.
An analysis in the vertical direction

should also be carried out on bridges that
have prestressed concrete superstructures
likely to be damaged by upward forces.
In both the analyses and design, the
effects of concurrently loading in more
than one direction may be neglected.
10.4 INPUT GROUND MOTIONS
The input ground motions used for
time-history analyses should satisfy the
following criteria;
(a) The records should contain at
least 15 seconds of strong ground
shaking or have a strong shaking
duration of 5 times the fundamental
period of the structure, whichever
is greater.
(b) The ordinates of the input ground
motion spectra should not be less
than 90% of the design spectrum
over the range of the first three
periods of vibration of the
structure in the direction being
considered.
The bridge should be analysed
using two different input motions for
each direction and the maximum computed
responses from the two inputs should
be adopted for design. The input
motions may be assumed to be in phase
at the base of all supports.
10.5 FOUNDATION INTERACTION
The influence of the soil foundation
on the response should be taken into account
using the recommendations given in Section
4.
10.6 MATERIAL PROPERTIES
10.6.1 Elastic Stiffness
Effective cracked section properties
should be used in concrete members where
forces arising from the design earthquake
loading exceed computed cracking
capacities.
10.6.2 Inelastic Moment Curvature
Idealizations
The moment curvature relationships
recommended for use in plastic hinge
regions of reinforced concrete, steel and
prestressed concrete members are shown in
Figs 10.1 and 10.2. The idealization
recommended for reinforced concrete and
steel members is a bilinear non-degrading
hysteresis loop with a 3% strain hardening
ratio. The recommended relationship for
prestressed concrete has been taken from
Blakeley and ParklO. 1,10. 2 j the non-
a n c
degrading loop shown is only applicable
before the onset of concrete crushing.
Further details including post crushing
loops are given in references 10.1 and 10.2.
10.6.3 Damping
The overall damping in the bridge
system expressed as percent of critical
equivalent viscous damping should generally
295
be taken as 5%. This value includes both
the structural damping and the damping
arising from radiation and inelastic
behaviour in the foundation. For special
structures such as long span steel cable
supported bridges that remain elastic
under earthquake loading, a lower value
of damping may be appropriate. Recommend-
ations for these bridges are given in
Section 11.
If the program allows individual modal
damping ratios to be input then a constant
value should be used for all modes. If
the program assumes Rayleigh type damping,
then the values of the coefficients a and
8 should be determined assuming the two
most dominant modes of vibration in the
initial elastic system to have the damping
values given above.
10.7 DESIGN FORCES AND DEFORMATIONS
10.7.1 Modal Spectral Analysis
The design forces and deformations
obtained from a modal spectral analysis
should be compared with the values
obtained from the simplified code analysis
method. It is unlikely that the two
methods will give results that differ
significantly but where differences
exceed 20% the reason for the variation
should be investigated.
10.7.2 Time-History Analysis
The overall ductility demands computed
by a time-history analysis should not be
greater than the available structural
displacement ductility factors. As
specified in Section 2.1.2, the overall
structural displacement ductility factor
should not exceed six and individual
member displacement ductilities should
not exceed eight unless the requirements
for additional ductility given in Section
5.2 are satisfied.
REFERENCES;
10.1 Blakeley, R.W.G., & Park, R. , (1973)
"Prestressed Concrete Sections with
Cyclic Flexure" Journal of the
Structural Division, A S C E , Vol 99,
No ST8, August 1973.
10.2 Park, R., and Blakeley, R.W.G., (1973)
"The Response of Prestressed Concrete
Structures to Earthquake Motions",
New Zealand Engineering Vol 2 8 ,
No 2, February 1973.
COMMENTARY:
General Comment -
Techniques are available to compute
the earthquake response of any bridge
structure and foundation soil model.
Recent refinements in numerical methods
and developments in structural mechanics
enable, at least in theory, an almost
exact representation of the bridge and
foundation. - '
0 1 0
- .
5
The main
c 1 0 6
limitations to obtaining an exact solution
are:
296

FIG. 10-1 MOMENT - CURVATURE IDEALIZATION FOR PLASTIC

HINGE REGIONS IN REINFORCED CONCRETE AND
STEEL MEMBERS

REGIONS IN PRESTRESSED CONCRETE


(a) Available design time and
resources.
(b) The accuracy to which most of the
basic input parameters can be defined.
Difficulty in specifying sufficiently
exact input parameters may arise in the
following areas:
(a) Detailed character and magnitude
of the earthquake ground motion.
(b) The difficulties in carrying out
sufficiently detailed, site invest-
igation so as to define the features
of the local geology that affect
the response of the structure.
(c) The inability to confidently
determine soil dynamic properties.
(d) Limitations in the knowledge of
the inelastic behaviour of
concrete members subjected to multi-
directional forces.
Assuming the resources were available,
it would be possible to carry out the
necessary investigation and testing to
define the input parameters mentioned
in areas (b) to (d) to an adequate degree
of precision for a reasonably exact
representation of a bridge and its
foundation system. However, it is unlikely
to be economic to obtain detailed data
in most of these areas, and this combined
with the uncertainty in the exact
character of ground motion on saturated
ralatively soft alluvium common at bridge
sites, or on the underlying bed rock,
severely restricts the accuracy that
can at present be achieved. Thus in many
practical situations the precision
provided by the computer analytical
method will greatly exceed the exactness
to which the input parameters can be
defined.
Many bridge structures have most of
their mass concentrated at a single level
and thus higher mode influences on the
dynamic response is generally of lesser
importance in bridges than buildings.
Generally bridge lateral load resisting
systems will have only a limited number
of well defined areas in which inelastic
action or plastic hinging is likely to
occur and this is a further factor that
generally enables bridges to be more
exactly represented than buildings by the
simplified code approach. However, factors
such as large lateral dimensions, expansion
joints, permanent ground movements and
soil-structure interaction may add
considerable complexity to the dynamic
behaviour of some bridges. These
factors are difficult to represent exactly
in a sophisticated analysis and thus
overall it is unlikely that dynamic
analyses will lead to a significant reduct-
ion in construction cost or improvement
to the design. Dynamic analyses can
however provide benefits in some cases
by giving a better insight into the
actual behaviour during earthquakes (e.g.
number of cycles of severe deformation)
and also by allowing a wide range of
297
poorly defined inputs to be examined.
At the present time it appears that
for a given design effort, greater bene-
fits can be achieved by devoting attention
to refinements in both structural form
and the detailing in the structural
members resisting the earthquake forces,
rather than by attempting to refine the
analysis. Thus it is recommended that
dynamic analysis should not be used for
the majority of bridge structures where
the dynamic behaviour can often be readily
predicted by simple analysis. Methods
of analysis involving greater complexity
than the code approach should be used
only for large and important structures
and where some feature of the design is
likely to add significant complexity to
the dynamic response.
CIO. 1
Major bridge structures of particular
importance would generally include bridges
located on major traffic routes of unusual
height or in excess of 200m long, and
having a high capital cost.
Where the mass of the pier is a
significant proportion of the total mass
contributing to inertia loading, as may
be the case for very tall piers, lateral
response of the piers as vertical beams
with distributed mass may cause an increase
in ductility demand at intended plastic
hinges, an increase in seismic shear
forces, or may cause plastic hinges to
form at locations (eg mid-height) other
than those intended by the designer. Such
behaviour cannot be readily predicted by
using an equivalent static lateral force
and a dynamic analysis may be necessary.
Hydrodynamic effects should be considered
for piers in deep water, as discussed in
Section 11.3.7. Articulation of the
superstructure by expansion joints or
hinges can result in deformation modes
that are difficult to calculate using
simple methods. The forces acting at the
joints may include friction, impact
and inelastic extension of seismic linkages.
Dynamic analyses may be necessary to obtain
a realistic estimate of relative displace-
ments . Unusual structures, including
suspension, cable stayed or arch bridges
do not conform to the simple single-degree-
of-freedom model implicit in code
equivalent lateral loading. This aspect
together with the generally high capital
cost of such structures will generally
make a dynamic analysis advisable.
CIO.2.1 If the lateral load resisting
system is expected to remain elastic under
the design earthquake then it is recommended
that the elastic modal spectral analysis
method of dynamic analysis be used in
preference to time-history numerical
integration or other methods. The modal
spectral analysis method requires signif-
icantly less computer time that other
methods and can be carried out by hand if
necessary. In the elastic modal spectral
analysis method an elastic system is
assumed for calculating the natural
frequencies and mode shapes and the maximum
response of each mode to the design earth-
298
quake is obtained from the code elastic
response spectrum. Because of the
simplifying assumptions, the time at
which each modal maximum response occurs
is not known and an approximate method
of combining the modal maxima, such as
the square root of the sum of the squares
method (SRSS), has to be used. Sufficiently
accurate solutions can usually be obtained
from the first two or three modes in each
direction.
The difficulty inherent in the
modal spectral approach of deciding how
the individual modal maximum responses
should be combined can be overcome by
carrying out an elastic time-history
analysis C I O . 2 1 The disadvantage of
b
this approach is that computation costs
are high and generally the improvement
in precision is small. For this reason
it is recommended that the modal spectral
method be used where the structure remains
essentially elastic.
CIO.2.2 The elastic modal spectral method
can be extended to analyse systems that
respond with moderate amounts of inelastic
behaviour (ductility factors 2 or less)
CIO. 7,CIO. 8 u • -,
' by assuming an equivalent
system. A simple approach is to use the
resonant amplitude matching (RAM) method
described by J e n n i n g s - ^
0 1 0
The
equivalent linear system has the same
mass as the yielding structure and a
period corresponding to the structures yield
period. The equivalent critical damping
coefficient is given by:
v Computer programmes are available for
2 1 2
C = — (1 - — ) (1 - + 0.05
}47T u
where u = overall structural displacement
ductility factor
k^,k^ = elastic and post yield stiffnesses
respectively of the idealized
bilinear system.
The above expression is plotted in Fig.
C. 10.1.
CIO.2.3 If a bridge is expected to have
significant inelastic response and is
not capable of close approximation by a
single-degree-of-freedom system then the
inelastic time-history method is likely
to be the most satisfactory approach.
Although a modal inelastic response
spectrum method can be used to give
an approximation to the earthquake
r e s p o n s e , • the main difficulty with
0 1 0 1
this method is that inelastic systems do
not have classical normal m o d e s , and even
where a modal representation may be an
acceptable approximation it is possible
that the inelastic behaviour will be
important in some modes and have little
influence on others. A simple example
illustrating this point arises when the
yielding member of a multispan bridge
is located at the abutment and significant
forces are produced by a beam type vibration
mode of a relatively tall massive pier.
CIO.3 Ideally a dynamic analysis should
be performed on a three dimensional
structural model with three orthogonal
earthquake input components. Bearing
in mind the high computational costs involved
with this type of analysis it is recommended
that a satisfactory approach is to consider
the three principal directions, two horizontal
and one vertical, to be acting independently.
At some distance from an earthquake
source the amplitude of ground shaking will
be strongly dependent on orientation with
respect to the s o u r c e ; • ! 5 thus assuming
0 1 0
that the design earthquake represents the
maximum amplitudes in any direction and
also allowing for the probability of peak
components of response occurring at the
same time, the combined total stresses will
not grossly exceed the maxima from the
individual components.
The lowest vertical modes of
vibration of many bridges have periods
that lie close to the peak of typical
earthquake response spectra and field
measurements have indicated low damping
values associated with these modes. Thus
quite high vertical accelerations may result
and should be taken into account in the
design of prestressed concrete superstructures
that could fail in a non-ductile manner when
subject to uplift forces.
CIO.4 Input ground motion accelerograms
used for dynamic analysis need to be consist-
ent with the design spectrum over the
frequency range of the modes of vibration
contributing significantly to the response.
Suitable ground motions can be obtained by
scaling recorded motions or by generating
consistent artificial accelerograms.
generating accelerograms that have pre-
described spectra °.10,C10.12,CIO.S7,C10.19
cl
and suitable records consistent with the
design spectra given in Section 2 are
being prepared.
Most bridges have span lengths of the
same order as the wave lengths of the
higher frequency waves that are associated
with the high accelerations peaks in the
ground motion. Thus it is clear that
there will be large phase differences
between the acceleration peaks at the var-
ious supports and this factor is likely to
result in a significant reduction of the
vibrational response that would be obtained
by assuming all supports move in p h a s e - 2 3 ^
0 1 0
The peak displacement in the recorded ground
motions are generally associated with
relatively long wave lengths and so
significant out of phase displacements
are only likely to need consideration for
bridges longer than 200m. Out of phase
movements resulting from elastic wave
propogation in the soil could result in
relative displacements of the order of
10mm. In addition, permanent ground
movements could result in even larger
relative displacements and thus it would
appear desirable to design longer bridges
for a total out of phase movement of about
20mm. In soft soils it would be prudent
to consider the effects of movements greater
than this value.
A number of research projects are
currently being conducted to study the
influence of out of phase inputs on the
 0-25 i—

1-0 2-0 3-0

Ductility factor JJ

F I G . C 10-1 EQUIVALENT VISCOUS DAMPING FOR BILINEAR

HYSTERETIC SYSTEM
300
vibrational response of bridges but at
the present time insufficient results
have been published to enable satisfactory
conclusions to be drawn. Inclusion of
out of phase inputs adds considerable
complexity to a dynamic analysis and thus
it is recommended that at present the
analysis should be simplified by neglecting
this effect.
CIO.6.1 For members intended to remain
elastic under design earthquake the
stiffness should be taken as the mean of
the value computed from the curvature in
the member at first yield of the tensile
reinforcement and the value equivalent
to the gross uncracked section of the mem-
ber. Design aids are available in
CDP 8 1 0 ° - / A for determining the
c l I 3
stiffness at first yield. For members
in which primary plastic hinging is
intended to occur, and a time-history
inelastic analysis is adopted, the
stiffness should be based on the curvature
a t first yield of the tensile reinforce-
ment. The equivalent stiffness for
moderate inelastic behaviour is covered
in CIO.2.2.
CIO .6.2 The hysteretic force-displacement
relationships recommended for prestressed
c o n c r e t e ^ • 3, CIO. 14 ^ reinforced concrete
c
and steel members are expected to be a
sufficiently accurate representations
for design purposes. The single most
important parameter required in an
inelastic analysis is the yield level
and since this can only be estimated to
a moderate degree of precision no increase
in accuracy would be obtained by using a
more refined force-displacement curve.
Under the design earthquake no significant
degradation should occur in the force-
displacement loop of a well detailed
plastic hinge or energy absorbing member
and thus a non-degrading loop has been
recommended.
CIO.6.3 In contrast with bridges, buildings
have non structural components and more
complex lateral load resisting systems
that may give rise to higher damping
values. Thus in general it is reasonable
to assume lower overall damping values for
bridges than normally assumed for buildings
and at present there appears to be no
justification for using overall values
higher than the recommended value of 5%.
Soil-structure interaction effects
may result in significant additional
damping on some bridge sites but at the
present time there is no conclusive
experimental evidence that would enable
specific damping values to be adopted in
design. Further research is continuing
in this area and eventually it should be
possible to make a rational allowance for
damping from soil-structure interaction.
If an analysis is carried out in which the
soil is modelled as an inelastic material
then the damping associated with the
structural system above the foundation
level should be taken as 3% see Section
4.3.
Wilson and P e n z i e n 1 0 • 2 2 have
c
described a method of setting up a
damping matrix to give specified values
of damping in each mode. The disadvantage
of this method is that it results in a
full matrix and greater computational
costs. However where the computer
program allows the complete matrix to
be set and running costs are not signif-
icant it is recommended that the Wilson
and Penzien method be used to give ident-
ical damping in each mode. Where Rayleigh
damping is assumed the damping matrix
is defined by
C = a M + $K
where M = the mass matrix
K = the stiffness matrix
a, 3 = scalar quantities that can
be evaluated by specifying
the damping and natural
frequency of any two modes •
For bridge structures, a satisfactory
damping matrix can generally be set up
by using the frequencies and damping of
the two most dominant modes of v i b r a t i o n - ^ •'
c
The damping specified in Section 10.6.3
should be used for the two modal values.
This procedure may result in very large
damping values being introduced in the
higher modes and an alternative approach
which reduces this effect is to specify
the damping in the first and a relatively
high mode, for example the tenth mode.
Although higher mode damping values have
been found to have a significant influence
on the response of building framesC10.4
it is unlikely that this would be the
case for bridge structures and either
of the above methods for setting a and
3 should be satisfactory.
CIO.7.1 Since the code design approach
is essentially a simplified modal
spectral analysis in general this method
should produce similar forces and
displacements to the code. Differences
between the two methods should only occur
when there is significant higher mode
response.
CIO.7.2 It is intended that if the designer
elects to carry out a time-history or
modal spectral analysis, that full
advantage should be taken of any benefits,
such as a reduction in strength, which
these more refined methods of analyses
may show to be j ustifiable.
CIO.8 REFERENCES:
Details of programs for dynamic
analysis are given in references C I O . 2 ,
CIO.9, C I O . 1 1 , CIO.18, CIO.20 and CIO.21.
Buckle^-- • gives a comprehensive set of
LU z
abstracts for bridge programs available
in New Zealand.
CIO.1 Anagnostopoulos , S.A., Haviland,
R.W. and Biggs, J.M., (1978) "Use
of Inelastic Spectra in Aseismic
Design", Journal of the Structural
Division , ASCE , Vol 104 , S T l ,
January 19 78.

CIO.2 Buckle, I.G., (1979) "Computer
Program Abstracts for Bridge
Analysis and Design", Report
217, School of Engineering,
University of Auckland, 1979.
CIO.3 Blakeley, R.W.G., and Park, R.,
(1973). "Prestressed Concrete
Sections with Cyclic Flexure",
Journal of the Structural
Division, ASCE, Vol 99, S T 8 ,
August 1973.
CIO.4 Carr A.J., Personal Communication,
University of Canterbury, Christ-
church.
CIO.5 Chen, M. and Penzien, J. (1975)
"Analytical Investigations of
Seismic Response of Short, Single,
or Multiple - Span Highway Bridges",
EERC 75-4, University of California,
Berkeley, January 1975.
CIO.6 Chen, M. and Penzien, J. (1977)
"Nonlinear Soil-Structure Inter-
action of Skew Highway Bridges",
EERC 77 2 4 , University of
California, Berkeley, August 1977.
CIO.7 Iwan, W.D. and Gates N.C. (1979)
"Estimating Earthquake Response
of Simple Hystertic Structures",
Journal Engineering Mechanics
Div., A S C E , Vol 105, No E M 3 ,
June 1979.
CIO.8 Jennings, P.C. (1968) "Equivalent
Viscous Damping for Yielding
Structures", Journal of the
Engineering Mechanics Division,
ASCE, Vol 94, No E M I , February,
1968.
CIO.9 Kanaan, A. and Powell, G.H. (1973)
"Drain - 2D General Purpose
Computer Program for Inelastic
Dynamic Response of Plane Structures,"
EERC 73-6, University of California,
Berkeley, April 1973.
CIO.10 Kaul M.K. (1978) "Spectrum -
Consistent Time - History
Generation", Journal of the
Engineering Mechanics Division,
ASCE, Vol 104, E M 4 , August 1978.
CIO.11 Mondkar, D.P. and Power, G.H. (1975)
"ANSR - 1, General Purpose Program
for Analysis of Nonlinear Structural
Response," EERC 75-37, University
of California, Berkeley, 1975.
CIO.12 Murakami, M. and Penzien, J. (1975)
"Nonlinear Response Spectra for
Probabilistic Seismic Design and
Damage Assessment of Reinforced
Concrete Structures," EERC 75-38
University of California, Berkeley,
November 1975.
CIO.13 N . Z . Ministry of Works and
Development, "Ductility of Bridges
with Reinforced Concrete Piers",
Civil Div. , Pub. CDP810/A, April
1975.
301
CIO.14 Park, R., and Blakeley, R.W.G.,
(1973). "The Response of
Prestressed Concrete Structures
to Earthquake Motions", New
Zealand Engineering, Vol 2 8 ,
No. 2, February 1973.
CIO.15 Penzien, J. and Watabe, M. (1975)
"Characteristics of 3 -
Dimensional Earthquake Ground
Motions", Earthquake Engineering
and Structural Dynamics, Vol 3,
1975.
CIO.16 Porter, F.L. and Powell, G.H.
(1971) "Static and Dynamic
Analysis of Inelastic Frame
Structures", EERC 71 - 3,
University of California,
Berkeley, June 1971.
CIO.17 Scanlan, R.H. and S a c h s , K. (1974)
"Earthquake Time Histories and
Response Spectra", Journal of
the Engineering Mechanics
Division, ASCE, Vol 100, No.EM4,
August 1974.
C10.18 Sharpe, R.D. (1974) "The
Seismic Response of Inelastic
Structures", Report N o . 74-13
Department of Civil Engineering,
University of Canterbury,
November 1974.
C10.19 Tsai, N.C. (1972) "Spectrum -
Compatible Motions for
Design Purposes", Journal of
the Engineering Mechanics
Division, ASCE , Vol. 98 ,
No. E M 2 , April 1972.
C10.20 Tseng, W . S . , and Penzien , J.
(1973) "Linear and Non-linear
Seismic Analysis Computer
Programs for Long Multiple -
Span Highway Bridges," EERC
73-20 , University of California
June 1973.
C10.21 Vanmarcke, E.H., Biggs J.M.,
Frank R., Gazetas , G. , Arnold,
P., Gasparini , D., and Luyties ,
W. (1976), "Evaluation of
Seismic Safety in Buildings,
Report No. 3, Comparison of
Seismic Analysis Procedures
for Elastic Multi-Degree
Systems, Publication N o . R76-5,
Department of Civil Engineering,
MIT, January 1976.
C10.22 Wilson, E.L., and Penzien, J.
(1972) "Evaluation of Orthogonal
Damping Matrices" International
Journal for Numerical Methods
in Engineering, Vol 4, January
1972.
C10.23 W u , R.W., Hussian, F .A. and
Liu L.K., (1978)"Seismic Response
Analysis of Structural System
Subjected to Multiple Support
Excitation", Nuclear Engineering
and Design 4 7 , 1978.
302

SEISMIC DESIGN OF BRIDGES

SECTION 11
BRIDGES REQUIRING SPECIAL STUDIES
M.J.N. Priestley*, P.R. Stanford**, A J . Carr***

11.1 GENERAL most modal techniques. Low damping

can be expected from elastic response
Some bridges, as a consequence of of continuous steel superstructures in
structural complexity, unusual ground the longitudinal and transverse directions
conditions, or high cost or risk potential, (11.1, 11.2, 11.3) _ i . .
merit special studies to better define For analysis it is
seismic performance. To a large extent recommended that a value of 2% of critical
these categories have already been damping be assumed.
mentioned in other sections of this
In the vertical direction, cable-
report on the seismic design of bridges.
stayed bridges may have high damping due
In particular, those bridges requiring
to unequal cable lengths, and the non-
consideration for dynamic analysis have
linear load-displacement characteristics
been defined in section 10. This section
of the bridge. If the displacements are
attempts to identify special problems
of small amplitude and response essentially
associated with specific situations and
linear, then this large damping may not
structural types, and where appropriate
be apparent. Vertical response of
makes recommendations about suitable
suspension bridges in the higher modes may
methods of analysis and detailing. The
have low damping, and should be considered.
approach adopted is deliberately less
formal than in previous sections, as the
There is little information available
intention is more to bring possible
on the ductility of such large elements
problems to the attention of the designer
as are commonly used for piers of cable-
than to present rigid design rules.
supported bridges. Consequently such
piers should be designed for low ductility
11.2 SPECIAL STRUCTURAL TYPES
levels, or if possible, to remain elastic
under the design-level earthquake. Because
Preceeding sections have concentrated
of large distances between major piers,
on common New Zealand bridge types -
large out-of-phase displacements may occur.
mainly structures supported on piers and
Although the superstructure flexibility is
abutments, where seismic response may be
likely to be such that these displacements
limited by plastic hinging of vertical
can be easily accommodated, checking is
elements, or simple spans either built
necessary. Analyses should be based on
monolithically with, or flexibly supported
a relative longitudinal or transverse dis-
by, abutments. A number of less common
placement of adjacent major piers equal to
structural types merit special consideration.
twice the maximum response displacement
A given by section 2 for the appropriate
11.2.1 Cable-supported Superstructures
seismic zone for a period of 3.0s. This
corresponds roughly with the maximum expected
Suspension and cable-stayed bridges
relative free-field ground displacement.
have structurally complex superstructures.
Because of the tupically long spans involved,
The high capital value of cable-
superstructures are relatively flexible
supported structures, and the severe conse-
both vertically and transversely. In
quences of major damage or collapse indicate
assessing transverse and longitudinal
that it will generally be appropriate to
seismic forces induced in superstructure
design for a low probability of occurrence
elements or transmitted to supporting
of the design earthquake within the expected
piers, realistic estimates of mode shapes,
useful life of the structure.
including consideration of flexibility
of the superstructure, must be adopted.
11.2.2 Arch Bridges
It is probable that as a result of high
fundamental periods, response will be
It is difficult to detail arch bridges
dominated by higher mode effects. As
for ductility, and where possible, they
the superstructure will be designed to
should be designed to respond elastically to
remain elastic during the design level
the design-level earthquake. A detailed
earthquake, analysis may consist of an
structural analysis will be necessary to
elastic modal analysis approach, with
define longitudinal and transverse mode shapes.
maximum response forces and displacements
Seismic forces and displacements may be
found from the square root of the sum of estimated using the SRSS method.
the squares (SRSS) of the individual modal
maxima. However, under large amplitudes
Special consideration must be given
of vibration geometric non-linearity
to relative longitudinal displacements of
effects in both suspension and cable-
the arch springing due to out-of-phase
stayed bridges will tend to invalidate
ground motion and seismic response displace-
ments of the typically steep embankments.
* Reader in Civil Engineering, University
Detailed geotechnical investigations should
of Canterbury, Christchurch
be carried out to establish the competence
** Ministry of Works and Development, Wgtn and stability of the embankments under seismic
*** Senior Lecturer in Civil Engineering, conditions.
University of Canterbury
B U L L E T I N OF T H E NEW Z E A L A N D N A T I O N A L SOCIETY FOR E A R T H Q U A K E E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980
 303

[ 1
to

Plastic hinge
locations

n—n

m ^ 7

(a) Single Column (b) Framed Pier

Fig 11-1 INERTIAL RESPONSE OF TALL SLENDER PIERS

24.5m 110m 24.5m

777?"

(a) Elevation

(b) T r a n s v e r s e Mode Shapes

Fig. 11-2 MODE SHAPES OF A MAJOR RAIL BRIDGE

304
Two-pin or encastre arches will
be less able to accommodate relative
longitudinal displacements than three-pin
arches.
11.2.3 Stress-Ribbon Bridges
It is unlikely that stress-ribbon
construction will be used in New Zealand
for other than minor structures, such
as footbridges. Although there is little
information available on the seismic
performance of stress-ribbon bridges, it
is expected that the inherent flexibility
should result in satisfactory response.
Horizontal forces and displacements should
be based on realistic estimates of trans-
verse and longitudinal mode shapes,
including consideration of the super-
structure flexibility.
11.3 Bridges With Extreme Geometry
Some bridges of conventional types
merit special studies as a consequence
of extreme height, length, width, skew or
curvature.
11.3.1 Bridges with Tall Piers
Tall piers may be subjected to high
inertia forces due to the response of the
distributed mass of the pier itself.
Ductility demands at intended plastic
hinge locations may be substantially
increased, and there may be a potential
for plastic hinging to occur at a location
close to mid-height, resulting in the
formation of a collapse mechanism (see
fig. 11-la). If preliminary analyses
based on elastic response of distributed
mass systems (e.g. ref. 11.4) indicate the
possibility of such behaviour, the only
realistic analysis will be a full dynamic
inelastic time-history analysis of the
pier/superstructure system. Tall piers
should be checked for additional moments
due to P-A effects. Results of such
an analysis may be highly dependent on
stiffness values assumed for the piers.
A detailed assessment of stiffness
variation up the column height should be
made, identifying probable cracked and
uncracked regions. Those regions sub-
jected to significant cracking or possible
yield should be assigned appropriate
cracked-section stiffnesses. It may be
necessary to carry out a sensitivity
analysis to establish the significance
of variation from assumed stiffness values
to overall response.
It may be found necessary to reduce
inertia response by using a framed pier
system as shown in fig.ll-lk longitudinally
and transversely.
If the superstructure of a bridge
in this category is restrained transversely
at abutments, bending of the superstructure
in the horizontal plane may be critical as
a result of large lateral displacements
of the pier tops. Similarly, for single
column piers, torsion of the superstructure
may be critical as a result of rotation
of the pier tops under lateral displace-
ments .
11.3.2 Bridges with Piers of Differing
Heights
When bridges span steep sided
river valleys, substantial differences in
adjacent pier heights may be inevitable.
Where possible, flexibilities of piers
should be adjusted to result in as uniform
yield displacements and ductility demands
on piers as possible. In some cases
this may be effected by adjusting the
flexibility of the supporting pile systems.
Where such attempts to 'regularise 1
the structural response are impractical,
analyses must establish the realistic
mass distribution to each pier, and how
this is influenced by sequential, rather
than simultaneous yielding of the
separate piers, particularly in the long-
itudinal direction. For multispan
bridges, transverse flexibility of the
superstructure will generally become
significant, particularly after initial
yielding of some piers. Lateral forces
transmitted back to abutments may be
substantially different from those
predicted by elastic analysis.
As an alternative to adjustment of
pier flexibility, it may in some cases
be easier to make all piers "stiff",
and use energy dissipating devices to
obtain ductility.
11.3.3 Long-Span Bridges
Transverse superstructure deformations
may become significant in major bridges
when the span/width ratio is large. This
will particularly be the case for major
rail bridges due to typically small widths.
Mode shapes and forces induced in the
support systems will be influenced by this
flexibility, and superstructure plastic
hinging may occur. Fig. 11-2 shows
theoretical transverse mode shapes
calculated for the North Rangitiki Rail
Bridge on the North Island Main Trunk.
Analysis should at least be on the
basis of a modal superposition approach,
and dynamic inelastic time-history
analysis is advisable to check ductility
levels.
Response of the superstructure to
vertical acceleration should be examined.
11.3.4 Highly Skewed Bridges
Examination of skewed bridges after
recent earthquakes has established a
tendency for rotation of the bridge
centreline about a vertical axis towards
the acute corners (see fig. 1 1 - 3 ) . In
the San Fernando earthquake of 1971 there
were examples where this caused severe
damage to columns, as a result of torsional
shear forces, and to abutments, as a result
of increased longitudinal displacements
and inadequate seating lengths^-!*5.
Although there is some theoretical evidence
to support this b e h a v i o u r a l ^ the
mechanisms are not clearly understood.
Consequently conservative assumptions
should be made regarding necessary
 2 3
(a) Elevation

k 1 M 1 k 3 M 2 k 2 v

'£—WWWV—©— >/VVV @—AAAAAAA—/

k ,k1 2 Stiffness of e n d b e a r i n g s «• p i e r s
k 3 Stiffness of joint bearings + linkages

(b) Longitudinal Response

(c) Transverse Displacements

F i g . 1 V / . SEISMIC RESPONSE OF BRIDGE W I T H MOVEMENT JOINT

Fig.11-5 CIRCULAR BRIDGE COLUMN WITH

INNER CONFINED CORE

306
clearances, seating lengths and lateral
restraint at abutments, and seismic
shear forces induced in intermediate piers.
To cope with torsional moments and
increased flexural ductility demand,
resulting from rotation about the vertical
axis, conservative detailing should be
specified, with an increase in the amount
of confining steel above code requirements.
It should be noted that this increase will
be required over the full height of the
columns, not just within potential
plastic hinge zones.
11.3.5 Bridges with Large Horizontal
Curvature
The seismic response of multispan
bridges with large horizontal curvature
and/or large horizontal deflection angle
is difficult to predict with accuracy,
particularly when the superstructure is
torsionally stiff as will generally be
the case to improve live-load distribution.
Significant axial seismic forces may be
induced in individual piers, even when
each pier consists of a single column.
Abutments may be subjected to high
force levels. Hinging will occur at
the top of single-column piers cast
monolithic with the superstructure, at
an early stage, even in the transverse
direction.
The direction of relative movements
or restraining forces at joints in the
superstructure will be uncertain, and
allowance for this should be made in
detailing.
Analysis of curved multispan bridges
ideally requires full 3-D inelastic
analysis. As this will rarely be practical
or even possible, conservative assumptions
regarding forces in piers and required
ductility demand must be made.
11.3.6 Long Bridges with Internal Movement
Joints
The response of long bridge
superstructures separated into two or
more sections by internal movement joints
can be difficult to predict. Out-of-
phase ground movements, as well as
structural differences between the
separate sections may cause substantial
relative movements across the movement
joints.
For longitudinal response it will
generally be conservative to assess
seismic forces and ductility demands on
columns by considering each section
independently. Maximum fessible relative
displacements at the movement joints
could be obtained by considering peak
displacements of the adjacent sections
to be out of phase. However, a more
realistic estimate of forces and relative
longitudinal displacements may be
obtained by modelling the movement joint
s a spring system representing combined
shear stiffness of bearings and axial
stiffness of seismic linkage bolts,
and analysing the coupled system. This
approach is illustrated schematically
in fig. ll-4b. An elastic modal super-
position analysis should be sufficient
to establish design forces and displace-
ments . Out of phase ground movements
at supports may add to relative displace-
ments and should be considered.
The movement joint will act as a
hinge for transverse seismic response.
The resulting loss of stiffness to the
superstructure may result in increased
displacements, and hence increased
ductility demand to piers on either side
of the joint (fig. l l - 4 c ) . Realistic
modelling of the articulation, of restrain-
ing forces at the joint provided by bearing
and linkage stiffness, and possibility of
impact associated with total joint closure,
is necessary to ensure adequate prediction
of transverse response. In some cases
it may be desirable to make transverse
response more regular by use of link spans,
hinged at each end between abutments and
main sections of the superstructure.
11.3.7 Piers in Deep Water
The response of piers in deep water
is affected by hydrodynamic mass of a
volume of water being forced to move with
the pier. A reasonable estimate of the
hydrodynamic added-mass is the mass of a
circular cylinder of water of diameter
equal to the pier width perpendicular to
the direction of motion. and length equal
to immersed depth-*-!' 7,11.8,11.9^ This
mass should be added to the pier mass when
considering the seismic response.
11.4 BRIDGES IN UNDESIRABLE LOCATIONS
Whenever possible, the siting of
bridges in locations where adverse ground
conditions significantly increase seismic
risk should be avoided. Such locations
include sites crossing or immediately
adjacent to an active fault, steep slopes
with potential instability under earth-
quake conditions, and sands with a
potential for liquefaction.
11.4.1 Sites Across or Near Active Faults
Bridges crossing or immediately
adjacent to active faults may be subjected
to large relative displacements of adjacent
piers or supports as a result of surface
fault expression. Although the
probability of such occurrence at a given
location during the design life of the
bridge will be very low, the possibility
should be considered in assessing a suit-
able structural type. A conservative
design, particularly in terms of dis-
placement capabilities should be adopted.
Design of piers should aim at providing the
maximum capacity possible, by use of
extra confinement of plastic hinge zones.
It will be advisable to provide an inner
confined core (fig. 11-5) capable of
supporting the structural dead weight
on the assumption that the outer flexural
confinement will have failed under an
extreme event. This has the added
advantage that under moderate, though not
catastrophic inelastic displacements, the
piers may be repaired by cutting out and
replacing the buckled outer layer of steel.

Relative merits of continuous vs.
simple-span construction should be
carefully evaluated. Although simple-
spans have the advantage of additional
flexibility in comparison to continuous
structures, difficulty will be experienced
in ensuring the spans do not drop from
supports. The additional redundancy
of continuous monolithic pier/super-
structure construction will tend to reduce
the probability of total collapse.
Accelerographs of recent earthquakes
indicate that vertical ground accelerations
close to a fault can substantially exceed
1.Og. This also supports the desirability
of continuous monolithic construction.
Additional pier flexibility may be
obtained by locating pile caps below
ground surface to increase pier length.
It should be recognised that the
purpose of design for such an extreme
event will be to avoid, or at lease minimise,
loss of life by reducing the probability of
total collapse. After such an earthquake
it is probable that the bridge would have
to be demolished and replaced.
11.4.2 Slopes with Instability Potential
Many bridges are inevitably sited
across steep-sided valleys. Detailed
geotechnical investigations should be made
to assess potential for slope instability
under seismic attack. For major structures
these investigations should include
geological and geomorphic studies including
expert study of aerial photographs, for
evidence of bank movement under recent
earthquakes, as well as material testing
and extensive bore-hole and trenching
investigations to check for unstable layers
and vertical fissures. Particular attent-
ion should be paid to drainage to prevent
infiltration of surface water and increased
porewater pressures in potential failure
regions. Special studies should be made
to investigate the practicality of improving
factors of safety against slope failure by
such means as unloading the banks by
removal of some top material. It may be
advisable to site each abutment well back
from the top of the slope, and tie back
any intermediate pile caps located on the
bank using rock anchors or other techniques.
11.4.3 Liquefiable Foundations
The assessment of liquefaction potent-
ial , and methods to reduce it, have been
covered in section 4.
There does not appear to be any
viable method to design a bridge to
remain serviceable if liquefaction occurs
under one or more pier foundations. As
liquefaction can occur at considerable
distances from the epicentre, the factor
of safety against liquefaction must there-
fore be high. In rare cases when the
factor of safety is felt to be only
marginally acceptable, design should aim
to provide maximum feasible ductility
capacity to avoid total catastrophic
collapse. The arguments of section 11.4.1
307
indicate that increased ductility detailing
and continuous monolithic construction
will generally be appropriate.
11.5 REFERENCES
11.1 Troitsky, M . S . , "Cable-Stayed
Bridges - Theory and Design",
Crosby Lockwood, Staples, London,
1977.
11.2 Selberg, A., "Dampening Effects
in Suspension Bridges", Pubis.
Int. Ass. Bridge Struct. Engng.,
10, pp. 183-198, 1950.
11.3 Hjorth-Hansen, E., Sigbjornsson, R.,
"Aerodynamic Stability of Box
Girders for the Proposed Stromstein
Bridge", Division of Structural
Mechanics, The Norwegian Institute
of Technology, University of
Trondheim, Norway, Feb. 1975.
11.4 Biggs, J.M., "Introduction to
Structural Dynamics", McGraw-Hill,
New York, 1964, 341pp.
11.5 Gates, J.H., "Factors Considered
in the Development of the California
Seismic Design Criteria for Bridges".
Applied Technology Council Workshop
on the Research Needs of Seismic
Problems Related to Bridges, San
Diego, 1979, 21pp.
11.6 Chen, M.C . and Penzien, J.,
"Soil Structures Interaction of
Short Highway Bridges", Applied
Technology Council Workshops on
the Research Needs of Seismic
Problems Related to Bridges, San
Diego, 1979, 33pp.
11.7 Newmark, N.M. and Rosenblueth,
E., "Fundamentals of Earthquake
Engineering", Prentice-Hall
International, N.J., 1971, 640pp.
11.8 Jacobsen, L.S., "Impulsive
Hydrodynamics of Fluid Inside a
Cylindrical Tank, and of Fluid
Surrounding a Cylindrical Pier",
Bull. Seism. Soc. Am. Vol. 39,
No. 3, 1949, pp.189-204.
11.9 Levy, Wilkinson, "The Component
Element Method in Dynamics",
McGraw-Hill, Chapter 9, 1976.
308

SEISMIC DESIGN OF BRIDGES

SECTION 12
STRENGTHENING OF BRIDGES FOR SEISMIC LOADS
J.F. McGuire*, R.W. Fisher**, P.R. Stanford***, P.J. North****

12. 1 INTRODUCTION the necessity for the region to

be reinstated quickly after a
Traditionally, bridge controlling severe earthquake.
authorities have not regarded the seismic
resistance of existing bridges as a (h) The need to facilitate movements
significant criterion, compared with capac- of emergency forces to and from
ity for live load. The preference is to the stricken region.
carry the risk, and face the possible
repair costs when the time comes. However, (i) The need to upgrade the strength
it is relevant to consider the seismic of the bridge for reasons other
risk in comparison with other risks and than seismic capacity; e.g. live
several techniques are available. Cost- loads , flood loads, alignment.
benefit analysis is one but it is admittedly
difficult to define criteria for analysis, 12.3 STRENGTHENING CRITERIA
since a number of the benefits are not
very easily measured in monetary terms. 12.3.1 Design Earthquake
It is easier to determine relative
priorities for strengthening among a The design earthquake for strength-
specific group of bridges, and methods ening should preferably be the same as
have been published for doing thisl2.1, that defined in Section 2, where the design
12.2,12.3^ life is taken as the remaining economic
life of the strengthened bridge, and the
It is not necessary to strengthen earthquake return period is determined
every bridge to full design criteria. as a function of the importance of the
There will be many cases where some bridge as provided for in Clause 1.3.5.
small attention to detail will significantly
improve the performance. Alternatively, if the basic
foundation elements are weak, and not
12.2 RELATIVE IMPORTANCE OF BRIDGES easily strengthened, it may be necessary
FOR STRENGTHENING to design for a smaller percentage of
the design load, on the basis that any
Points which should be considered in improvement will still be an advantage.
evaluating the importance of a bridge
(refer to Clause 1.3) and in setting 12.3.2 Structural Design
priorities for strengthening include:
As far as is practical the
(a) The existence of an alternative principles of capacity design should
route. be followed in accordance with Section 3.
Seismic details should be appropriate to
(b) The existence of services carried the load level chosen for the strengthening
by or passing under the bridge. and be in accordance with Clauses 12.3.3
(a) to (d) and other relevant Sections of
(c) The existence of buildings under this document.
or nearby which could be damaged
in the event of a collapse. Weak points of the structure should
be identified and strengthened if this is
(d) The expected remaining economic technically and economically feasible.
life of the bridge. Designers should be aware that the strength-
ening of one or even several elements of
(e) The cost and the benefit derived a structure will not necessarily improve
from strengthening to various its overall capacity to resist an
proportions of the design loading. earthquake if other critical elements or
details remain inadequate. A thorough
(f) The availability of suitable tem- understanding of the structure is necessary
porary bridging or modes of for adequate design of the strengthening.
transport not requiring bridging.
12.3.3 Design Details
(g) The likely needs of the region and
The range of techniques for strength-
ening is wide * . The appropriate
* Senior Engineer, Ministry of Works and method of strengthening will depend on the
Development, Wellington specific requirements for the bridge
** Assistant Bridge Engineer, Chief which should be determined after adequate
Civil Engineers' Office, NZR, Wellington site inspection, review of design
*** Senior Design Engineer, M.W.D., Wellington calculations, construction drawings and
**** Consulting Engineer, Hamilton analysis. In some cases dynamic analysis
may be desirable. (Refer to Section 10).

B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR EARTHQUAKE E N G I N E E R I N G , V O L 13 NO. 3 SEPTEMBER, 1980

 309
If a large number of bridges have to be
evaluated in a short time it may be
necessary to group the different bridge
types and develop standard solutions for
each type.
Specific measures which will
generally lead to significant enhancement
of seismic resistance are:
(a) Improvement to linkage bolts or
cables and tiebacks and seating
at joints in the superstructure
and renewal of sliding bearings
which may have ceased to function.
Renewal of bearings may improve
the seismic response of the
structure and control the
distribution of loads between
supports.

(b) Improvement to the ductility of

the substructure by the addition
of suitable confining reinforcement
or the addition of energy absorbing
devices.

(c) Improvement to the strength of

the substructure and foundations.
Weak piers or foundations may be
improved by the provision of
carefully located prestressed tendons.

(d) Improvements to the stability of

approaches.
12.4 REFERENCES:
12.1 Longinow, A., Cooper, J.D., Bergmann,
E., "Selection of Critical Bridges
for Retrofitting", Journal of
ASCE Vol 105 No. TCI, April 1979.

12.2 Longinow, A., Robinson, R.R., Chu,

K.H., "Retrofitting of Existing
Highway Bridges Subjected to
Seismic Loading - Analytical
Considerations", ATC-6 , paper lib.

12.3 Ohashi, M., Fijii, T., Kuribayashi ,

E., Tazaki, T., "Inspection and
Retrofitting of Earthquake
Resistant Vulnerability of Highway
Bridges - Japanese Approach",
ATC-6, paper 11c.

12.4 Dengenkolb , O.H., "Retrofitting

of Existing Highway Bridges
Subject to Seismic Loading -
Practical Considerations", ATC-6,
paper 11a.
 3

FURTHER COMMENTS ON SEISMIC DESIGN LOADS FOR BRIDGES

J.B. Berrill*, M.J.N. Priestley* apd R. Peek**

SUMMARY:

This paper provides background material to the loadings section

of the model code recently published by the Society's Discussion Group
on Bridge Design, and presents a preliminary re-evaluation of the design
spectra given in the proposed code. The basis for the proposed zoning
scheme, in which the present uniform Zone B is replaced by a transition
zone, is discussed. Arguments are given underlying the return period
coefficients, and the force reduction factor used in generating the
inelastic response spectra of the code. It is likely that the design
spectra and the values of the other coefficients determining base shear
forces will need to be revised as further research results become available;
however, the form of the base shear expression, and the loadings section
as a whole, should remain unchanged. Re-evaluated spectra suggest that
the seismic coefficient values given in the proposed code may be too large
by about 25 percent in Zone A, and too low by as much as 40 percent in
Zone C. While the reassessed values should be more reliable than the
original ones, they are based on a Japanese attenuation model, which has
not yet been calibrated against New Zealand data. Further research is
required to establish an appropriate attenuation model for New Zealand;
to avoid undue proliferation of design loadings it is preferable to defer
revision of the various coefficients in the proposed code until such a
model is available. Until this is done, the proposed spectra should be
viewed with caution, particularly in Zone C.

INTRODUCTION: SEISMIC BASE SHEAR FORCE:

A recent edition of the Bulletin The expression adopted for determining

(Volume 13, No. 3, September 1980) carried the design value of the horizontal seismic
the collected papers of a Discussion Group base shear force, H, was as follows:
on Bridge Design, sponsored by the New
Zealand National Society for Earthquake
Engineering. Of the various topics H = C H p Z H Mg (1)
covered by the discussion group, design
ground motions was one of the more where
difficult because contemporary seismic
risk analysis techniques had not been C„ = basic horizontal force coefficient
applied to New Zealand. The group was ti]l
faced with the alternatives of mounting Z n = return period coefficient
a major research program or of using ri
standard, but less precise, techniques.
In the event, the three-man loadings M = mass participating in lateral motion,
sub-committee was instructed by the group
as a whole to produce results quickly, and
using simple methods. This was done,
at the cost of some conservatism, in the g = acceleration due to gravity
expectation that the design spectra and
zoning map would be updated as results By introducing earthquake return
of more detailed studies became available. period into the determination of design
base shear force,- the risk of the design
The proceedings of the discussion value being exceeded (not necessarily
group were written in model code and equal to the risk of failure) is more
commentary form which limited the amount obvious, and is given some perspective.
of background material that could be Also, it allows the designer more freedom
presented. These notes are intended to in selecting an appropriate degree of
present some of the arguments underlying seismic resistance, and opens the way to
the loadings section of the group's cost-benefit studies and formulation of
proceedings, "Design Loading and Ductility a reliability-based design code*. It
Demand"1, and to present a preliminary was the intuitive feeling of a majority
reevaluation of the design spectra. of the discussion group that most bridges
should be designed for ground motions with
a 150 year return period. Hence C ^ R

*Department of Civil Engineering, University and Z were normalized so that Z = 1

TT
of Canterbury.
**Formerly, University of Canterbury; present for a return period of 150 years.
address : Department of Civil Engineering, *The discussion group considered seismic
M.I.T. effects only. A probabalistic design code
should also consider random variations in
dead, live and other loads, as well as in
structural strength, to arrive at risk factors
related to the o v e r a l ^ p r o b a b i l i t y of failure.
See, for example, Elms .
B U L L E T I N OF THE NEW Z E A L A N D N A T I O N A L SOCIETY FOR E A R T H Q U A K E E N G I N E E R I N G , V O L . 14, NO. 1 , M A R C H 1981
4
BASIC FORCE COEFFICIENT, C„
riU
The horizontal force coefficient,
C ^ , depends on the seismic zone, the
H shield the Waikato to some extent from
fundamental natural period of vibration
of the structure, and on the design value
of the displacement ductility factor, y.
For u = 1, C estimates elastic, absolute
TT
Hu
acceleration response, for 5 percent of
critical damping, with a return period
of 150 years. Estimates of inelastic
response are obtained by multiplying
elastic response values by a force
reduction factor. The derivation of the
elastic response spectra, the seismic
zones and the force reduction factor
are discussed below.
Seismic Zones -
The proposed zoning scheme, shown
in Figure 1, does not differ greatly
from that of the current New Zealand
design code, NZS 4 2 0 3 .
3
Three zones
are retained. The main change is to
Zone B, which in the proposed scheme
provides a smooth transition between
the two zones of assumed uniform seismicity,
A and C, reflecting seismic risk more
accurately than in the present scheme.
The map shown in^Figure 1 is based
principally on Smith's contour maps of
Modified Mercalli intensities for given
return periods, reflecting seismicity
observed during European settlement, and
on Lensen's "Late Quaternary Tectonic
Map" , indicating regions of active
faulting in geologically recent time.
In some areas, these sources are in conflict;
this is not surprising, since they result
from very different methods, and observe
short and long time periods respectively.
One area where recorded seismicity
seems lower than expected from geologic
evidence is in the Alpine Fault region
of the South Island. On the basis of
Smith's intensity contours, this area
should be in Zone B. However, it would
clearly be imprudent to exclude the
principal fault in the country from Zone A.
This conclusion is reinforced by Adam's
recent evidence^ for the occurrence of
great earthquakes on the Alpine Fault
at intervals of about 500 years in the past.
Taranaki is another difficult region.
High historical seismicity in this area
is reflected in S m i t h s maps, which suggest
1
that the Zone A - Zone B boundary should
lie in an east-west direction. But the
main tectonic features of the North Island
lie in a north east - south west direction.
These features comprise the Hikurangi trench,
and the^main mountain ranges and fault
systems . They indicate that the zone
boundary, as well as the Zone B risk
contours, should lie parallel to the east
coast of the North Island. However,
there is other geologic evidence to support
a more east-west trending boundary.
Earthquakes occurring in the region of
crustal tension north of Taupo, indicated
by the thermal area extending into the
Bay of Plenty and by the Hauraki grabben
(Lensen, personal communication), should be
less energetic than those occurring further
south at similar distances from the main
fault system. Also, the soft crustal
material of the thermal region should
seismic waves originating in the main
tectonic region. Hence we should expect
seismic risk in Taranaki to be higher
than in Waikato.
Other information used in determining
the boundaries and contours in Figure 1
were Matuschka's^ contour maps of intensities
and of peak ground accelerations and
velocities for selected return periods,
a map of epicenters of large earthquakes,
shown in Figure 2, and a map of density
of energy release, shown in Figure 3.
The latter two maps were prepared from
seismological data recorded from 1840
(Smith ).
9
These diverse and not particularly
consistent sources of information suggested
that a zone of uniformly high risk should
be retained in roughly the same region
as Zone A of the present standard. Also,
to provide a minimum level of seismic
resistance, a uniform Zone of low risk
should also be kept in the present North
Island Zone C. Elastic response spectra
for Zones A and C were computed from
Smith's^ 150 year M.M. intensities for
Wellington and Auckland respectively.
These locations were chosen because
they are in the more active parts of their
respective zones, and also because they
are the locations likely to have the
greatest amounts of construction. The
zone boundaries and contours were then
fixed, quite subjectively, by sketching
with the above points in mind and with the
various maps in view.
Elastic Response Spectra -
The elastic acceleration response
spectra from which the base shear force
coefficients were obtained were computed
from Smith's 150 year return-period
M.M. intensities. First, peak ground
accelerations and velocities were computed
from intensities, using the correlations
of Trifunac and Brady- and of Murphy
1,0
and O'Brienll. These were found to agree,
fairly closely with each other, and to be
consistent with values estimated independ-
ently by Matuschka . Five percent damped
elastic response spectra were then computed
from the peak motion values, again using
two methods, that of Seed et a l . and that
1 2
of Newmark and H a l l 1 3
and M o h r a z ^ .
1
Again,
the spectra found by the two methods agreed
quite closely. In Zone A the spectrum
derived by S e e d s procedure, for deep
1
cohesionless soil deposits, was used. In
Zone C M o h r a z s method was used, but with a
1
higher peak ground velocity than expected
from the 150 year Auckland intensity, to
allow, at least qualitatively, for the
effect of distant earthquakes.
The Zone B spectrum has the average
shape of the Zones A and C spectra, and
is normalized so that the steps in spectral
ordinates that occur in passing out of
Zone B are minimized, to a maximum value
of about 15 percent.
 KILOMETRES

Figure 1 - Proposed Seismic Zones. Figure 2 — Distribution of Earthquake Epicentres in New Zealand, Note that
Shorter Sampling Periods have been Used for Smaller Magnitudes.
6

Figure 3 - Map Showing Density of Seismic Energy Release.

PERIOD (sec)

Figure 4 — Proposed Zone A Acceleration Response Spectra, Compared with

Present M.W.D. Bridge Design Spectra and with Various Recorded
Strong Motions. Note that Elastic Acceleration Response {with JJ=1 )
Corresponds to CHJJ with u = 1 .

In Figure 4, the proposed Zone A
design spectra (or values of C__ ) are
ti]i
compared with corresponding Ministry of
Works and Development design spectrum for
b r i d g e s ^ and with some spectra of well-
known accelerograms. The proposed elastic
spectrum should be compared with six times
the M.W.D. curve since the accompanying
design procedure is based on a ductility
factor of six. It can be seen that the
two curves agree closely in the middle
period band. At short periods the proposed
curves result in much higher design loads
for ductile structures than given by the
existing spectra-^. At longer periods,
the two curves separate due to the artific-
ially horizontal plateau of the present
spectrum (a constant acceleration response
line implies unbounded relative displacement
response, which is not possible).
Note that the proposed Zone A elastic
spectrum is similar to that of the 1966
Parkfield, California accelerogram, recorded
close to the source of a small, M= 5.6,
earthquake. About four such earthquakes
occur per year in New Zealand.
Another point to note from Figure 4
is the divergence of the proposed, u = 6
spectrum from the present spectrum at
periods smaller than about 0.6 seconds.
This occurs because in obtaining the present
curves, a constant force reduction factor,
proportional to 1/y, was applied to
S k i n n e r ' s ^ average spectrum (also shown
in Figure 4 ) . However, as explained
below, use of a constant reduction factor
is not valid at periods smaller than about
0.7 seconds; because of this, the present
spectra are unconservative at small periods.
A similar comparison is made in Figure
5 for the Zone C spectra. Here, the
proposed values of C__ are much smaller than
the present ones. They are also smaller
than those from a more rigorous, but
preliminary, re-evaluation of the design
spectra which is discussed further below.
Return Period Coefficient Z__ -
1
rl
Values of Z n (Table 2.1) are
proportional to peak ground velocities,
computed from Smith's^ M.M. intensities
for various return periods. Trifunac
and B r a d y ' s
1 0
relation between peak velocity
and intensity was used. Originally,
separate values of Z„ were computed for
each zone, but they were so close that a
single set of values was used in the final
version of Table 2.1. Peak ground velocity
was used for scaling because it character-
izes spectral strength at around 1 second,
and is therefore more representative of
the strength of the spectrum as a whole
than, say, peak acceleration.
Site Conditions -
Site soil conditions are not required
to be considered in obtaining values of the
seismic coefficient C from the proposed
code. The single set ^ of curves giving
values of C ^ was estimated for deep
alluvial soils, typical of many New Zealand
bridge sites. It was felt that techniques
7
for predicting "site effects" were not
well enough advanced for precise
predictions of local effects to be made
reliably. Two exceptions are mentioned in
the commentary. They are hard rock
sites, where it is recommended that values
of C„ be increased at periods less than
nu
0.4 seconds, and sites with very soft,
uniform surficial deposits. The latter
is the only case in which it was considered
that one-dimensional "shear-beam" analyses
could be employed usefully. The problem
of assessing site effects, together with
other sources of uncertainty in estimating
strong ground motions, is discussed briefly
in reference 24.
Force Reduction Factor -
As mentioned above, the existing
loadings code for buildings (NZS 4 2 0 3 ) 6
and also the Highway Bridge Design B r i e f 2
have seismic coefficients established
by reducing an assumed elastic response
spectrum by a constant multiplier, R,
equal to the inverse of the displacement
ductility factor. That is,
This practice is based on the equal displace-
ment principle, which has been shown by
numerous inelastic time-history analyses
of simple oscillators to give a reasonable
estimate of the response of structures
with periods longer than about 0.7s. It
has, however, long been recognised that for
short period structures, the approach is
non-conservative, and could result in
excessive ductility demand.
The equal energy principle has been
proposed as being applicable for short
to medium period structures, and results
in a force reduction factor of
R =
— (3)
J2y t - 1
Thus for a displacement ductility factor
Of y " 6, the equal energy principle results
in a reduction factor of 0.302, compared
with 0.17 implied by the equal displacement
principle.
For very short period structures
Csay T <, 0.2s) the force reduction factor
given by Eq. 3 has been found to be still
unconservative. Gulkan and S o z e n
2 2
quote
displacement ductility factors of 28 to 30
resulting from a T - 0.15s structure designed
for a force reduction factor of about 0.3,
and analysed under different earthquake
records. The equal displacement and equal
energy principles would result in expected
ductility demand of 3.33 and 5.95
respectively. This inadequacy of the equal
energy principle for short period structures
results from a tendency for the period to
lengthen and degrade into a period-range of
higher response, as a result of inelastic
action^. For medium and long period
structures, the period lengthening causes
a shift away from the period range of
maximum response.
8

PERIOD (sec)
Figure 5 — Comparison Between Proposed Values of Seismic Coefficient CHu for
Zone C, and Those Obtained in a Preliminary Re-evaluation Using more
Rigorous Seismic Risk Estimation Techniques.

Figure 6 — Comparison Between Zone A elastic Acceleration Response (CHu for J J = 1 )

in Proposed Code, and Corresponding Spectrum Obtained in Preliminary Re-
evaluation using Peek's Seismicity Model and Katayama's Attenuation Model.

In the limit, when the period
approaches T = 0, the structural response
acceleration will be equal to the peak
ground acceleration, regardless of the
ductility factor. Consequently the force
reduction factor should be R = 1.0 at
T = 0.
The approach adopted in the loadings
section of the proposed bridge design
code was to reflect the above trends
by accepting the equal displacement
principle for T > 0.7s, use the equal
acceleration principle for T = 0, and to
specify a linear variation between these
methods for intermediate periods. Thus
the force reduction factors are for T = 0.
for T = 0 : R = 1.0
0.70
"for 0 < T < 0.7s : R =
(ji-l)T+0.7
1
for T ^ 0.7s : R
For ]i = 6, Eq. 4b gives the same
force reduction factor as the equal
energy principle for a period of T - 0.32s.
At lower periods, it is more conservative.
Although Eq. 4b is more the result of
intuition than extensive calibration, it
thus appears to provide a reasonable
estimate of response throughout the period
range.
PRELIMINARY RE-EVALUATION OF SPECTRA:
Application of probabalistic
seismic risk analysis techniques, pioneered
by C o r n e l l ^ to New Zealand has been
1
r
studied by Peek **.
1
He has established
a seismicity model for the country, based
on seismological, geologic and tectonic
data. This is one of the two principal
components of a seismic risk analysis,
and describes the distribution of earthquake
magnitudes as a function of geographical
position. The other component is an
attenuation expression, relating magnitude
to strength of shaking at a site, through
epicentral or focal distance. Peek
examined several attenuation expressions
available in the literature, and found
that none was immediately applicable to
New Zealand. Because of the high
sensitivity of predicted design, or
risk spectra to details of the attenuation
expression, worldwide and New Zealand data
are being studied in an attempt to derive
an attenuation model more representative of
New Zealand geologic conditions, and it is
questionable whether the spectra presented
in the proposed code should be revised
before results of this study are available.
However, preliminary estimates of
150-year response spectra have been made
using an attenuation expression derived
from Japanese strong motion accelerograms
by Katayama et al. ^ The Japanese
1
expression is favoured because of the
tectonic similarities between Japan and
New Zealand.
In computing the spectra for Auckland
and Wellington, shown in Figures 5 and 6
respectively, some judgement has been
9
exercised in selecting values of the ^
scatter parameter, in P e e k s notation
1
This parameter measures the spread of
spectral ordinates, treated as random
variables, about the mean attenuation curve.
(The greater the scatter, the greater the
probability of a given strength of ground
motion; hence the greater the spectral
ordinates for a given return p e r i o d .)
1
Katayama et al. suggest a value o f c T ^ = 0.3
Q
at all periods, and this would be the
appropriate value to use for sites in Japan.
However, unless New Zealand conditions are
identical to those of Japan, a greater
value of l
required.
s
Guided by
20
Trifunac and Anderson , values of
(4a) increasing from 0.3 at a natural period
of 0.2 seconds to 0.48 at 3 seconds have
been used.
(4b) Figures 5 and 6 suggest that the
values of C„ given in the draft bridge
(4c) 1
code are overestimated in Zone A by about
25 to 30 percent, and under-estimated in
Zone C by about 40 percent. However, the
spectral shapes are similar; bearing in
mind the uncertainties involved in both
methods of computations, the differences
are not unexpected nor particularly great.
However, they suggest that if use is to be
made of the draft code in Zones B and C
before a proper revaluation has been
carried out, higher values of C ^ should
H
be used. Above all, the comparisons
point up the importance of the attenuation
study in progress.
Re-evaluation of the return period
coefficient Z yields the results shown
0
n
in Figure 7. These suggest that the values
given in Table 2.1 of the proposed c o d e 1
are too large for return periods greater
than 150 years and too small for shorter
return periods. The close agreement between
the Auckland and Wellington values confirms
the use of a single set of coefficients
for all zones. As with the spectra
themselves, values of Z depend on the
H
attenuation expression used; again, it
is doubtful whether the values of Z^
in the proposed code should be revised
before the attentuation study in progress
is completed.
CONCLUSIONS:
The design spectra presented in the ^
recent paper on Bridge Earthquake loadings ,
and described further in this paper,
represent a first attempt to provide a more
rational basis for seismic design loading
for New Zealand bridges than was previously
available. Although the data for
establishing the spectra are sparse, and
seismic risk models for New Zealand are
still in preliminary stages of development,
the methodology presented is felt to be
sound. Consequently, improvements in
seismicity models and attenuation expressions
are expected to result in (hopefully small)
changes to the basic response spectra, but
should not affect the method of calculating
design loads. Already recent work has
indicated that the response ordinates for
10

Figure 7 — Values of Return Period Coefficient ZH Obtained in Preliminary Re-evaluation,

at Natural Periods of 0.2 and 3.0 sec, for Auckland and Wellington Sites. Values
from Table 2.1 of Proposed Code , are also shown.
1

Zone C may be low by as much as 40 percent.
It is expected that continuing research
will result in revised spectra that can be
adopted with an acceptable degree of
confidence within the next few years.
Until such time, the curves should be
viewed with some caution, particularly
for Zone C, though it is felt that they
represent a greatly improved option to the
existing spectra incorporated in New
Zealand bridge and building codes.
ACKNOWLEDGEMENTS:
The authors wish to acknowledge many
helpful conversations with members of the
Discussion Group, particularly with
H.E. Chapman, and with A.J. Carr and
R. Park, who reviewed the manuscript.
Thanks are also due to New Zealand
Electricity for financial support of the
seismic risk study used in the preliminary
re-evaluation.
REFERENCES:
1. Berrill, J.B., M.J.N. Priestley and
H.E. Chapman, "Design Earthquake
Loading and Due ti1ity Demand",
Bulletin N.Z. National Society
Earthquake Engineering, Vol. 13,
1980, pp. 232-241.
2. Elms, D.G., "Reliability-Based Risk
Factors", Bulletin N . Z . National
Society Earthquake Engineering,
Vol. 13, 1980, pp. 3-13.
3. "N.Z. Standard Code of Practice for
General Design and Design Loadings
for Buildings", NZS 4203 : 1976,
Standards Association of N . Z . 1976,
8 0 pp.
4. Smith, W.D., "Statistical Likelihood
of Earthquake Shaking Throughout
New Zealand", Bulletin N . Z . National
Society Earthquake Engineering,
Vol. 9, 1976, pp. 213-221.
5. Lensen, G., "Late Quarternary
Tectonic Map of New Zealand", N.Z.G.S.
Misc. Map Series, No. 12, D.S.I.R.,
Wellington, 1977.
6. Adams, J., "Paleoseismicity of the
Alpine Fault Seismic Gap, New
Zealand", Geology, Vol. 8, 1980,
pp. 72-76.
7. Walcott, R.I., "New Zealand Earthquakes
and Plate Tectonic Theory", Bulletin
N . Z . National Society Earthquake
Engineering, Vol. 12, 1979, pp. 87-9 3.
8. Matuschka, T., "Assessment of Seismic
Hazards in New Zealand", Report No. 2 23,
School of Engineering, University of
Auckland, 1980, 214 pp.
9. Smith, W . D . , "The "N.Z. Quakes'
Package of Computer Programs",
Tech. Note No. 76, Geophysics Division,
D.S.I.R., Wellington, 1979.
10. Trifunac, M . D . and A.G. Brady,
"On the Correlation of Seismic
Intensity Scales with the Peaks of
Recorded Strong Ground Motions",
11
Bulletin Seism. Society Am.,
Vol. 65, 1975, 139-162.
11. Murphy, J.R. and L.J. O'Brien,
"The Correlation of Peak Ground
Acceleration Amplitude with Seismic
Intensity and Other Physical
Parameters", Bulletin Seism. Society
Am., Vol. 67, 1977, pp. 877-915.
12. Seed, H.B., C. Ugas and J. Lysmer,
"Site-Dependent Spectra for Earthquake-
Resistant Design", Bulletin Seism.
Society Am., Vol. 66, 1976, pp. 221-243.
13. Newmark, N.M. and W.J. Hall,
"Procedures and Criteria for Earthquake
Resistant Design", in, Building
Practices for Disaster Mitigation,
National Bureau of Standards, Boulder
Colorado, 1973, pp. 209-236.
14. Mohraz, B., "A Study of Earthquake
Response Spectra for Different
Geological Conditions", Bull. Seism.
Society Am., Vol. 66, 1976, pp.
915-935.
15. "Earthquake Resistant Design of
Bridges", N . Z . M.W.D. Draft
Publication C D . P . 707/A, 1978.
16. Skinner, R.I., "Earthquake-Generated
Forces and Movements in Tall
Buildings", Bulletin 166, N . Z . D.S.I.R.,
1964.
17. Cornell, C.A., "Engineering Seismic
Risk Analysis", Bull. Seism. Soc.
Am., Vol. 58, pp. 1583-1601, 1968.
18. Peek, R., "Estimation of Seismic
Risk for New Zealand : A Seismicity
Model and Preliminary Design Spectra",
Research Report No. 8 0-21, Dept.
of Civil Engineering, University of
Canterbury, 84 pp. 1980.
19. Katayama, T., Iwasaki, T. and Seaiki, M. ,
"Statistical Analysis of Earthquake
Acceleration Response Spectra",
Trans. Japanese Society Civ. Eng.,
Vol. 10, pp. 311-313, 1978.
20. Trifunac, M . D . and J.G. Anderson,
"Methods for Prediction of Strong
Earthquake Ground Motion", Report
No. NuREG/C12-0689, Dept. of Civil
Eng. Univ. of S. California, Los
Angeles, 1979.
21. "Highway Bridge Design Brief",
M.W.D. Civil Division Publication
CDP 701/D, Wellington, Sept. 1978,
52 pp.
22. Gulkan, P. and Sozen, M.A., "Inelastic
Responses of Reinforced Concrete
Structures to Earthquake Motions.",
Jour. A.C.I., Proc. V. 71, No. 12,
Dec. 1974, pp. 604-610.
23. Priestley, M.J.N., Park, R., and
Ng. K.H., "Influence of Foundation
Compliance on the Seismic Response
of Bridge Piers", Bulletin N . Z .
N.Z.N.S.E.E. Vol. 12, No. 1, March
1979, pp. 22-34.
24. Berrill, J.B., "Properties of Strong Ground
Motions", Proc. Seminar on Bridge Design and
Research, N . Z . Nat. Roads Board, Auckland,
Vol. 4, 1978, pp. 35-40.