Numerical

Modelling and Capacity Design of Earthquake-Resistant Reinforced Concrete Walls

Peter Linde

August

1993

Bericht Nr. 200

Birkhauser

Verlag

Basel

Boston

Berlin

Institut fr Baustatik und Konstruktion, ETH Zrich

A CIP

Library

catalogue record for this book is available from of Congress, Washington D.C, USA

the

Linde, Peter:
Numerical

modelling

and

capacity design of earthquake-

resistant reinforced concrete walls / Peter Linde. Institut fr

Baustatik und Konstruktion, Hochschule

Eidgenssische

Technische
:

(ETH),
1993

Zrich.

Basel ; Boston ; Berlin

Birkhauser,
(Bericht 200)

/ Institut fr Baustatik und Konstruktion, ETH Zrich ; Nr.

ISBN 3-7643-2968-8 NE: Institut fr Bauslatik und Konstruktion <Zrich>: Bericht

material is subject to Copyright. All rights are reserved, whether the whole or part of the concerned, spedfieally the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, kind of use permission of reproduction on microfilms or in other ways, and storage in data banks. For any This work is
the

Copyright

owner must

be obtained.

1993 Birkhauser

Verlag Basel,

P.O. Box 133, CH-4010 Basel, Switzerland

Printed

on

aeid-free paper

Printed in Switzerland ISBN 3-7643-2968-8

ISBN 0-8176-2968-8

98765432 1

Modelling and Capacity Design of Earthquake-Resistant Reinforced Concrete Walls

Numerical

Peter linde

Institut fr Baustatik und Konstruktion

Eidgenssische Technische

Hochschule (ETH) Zrich

Ju 1993

PREFACE

Reinforced concrete structural walls

are

very efficient elements for

protecting buildings

against excessive early damage and against collapse under earthquake actions. Structural
walls

designed by the

world-wide

increasingly used capacity design method exhibit design ductility according to well known rules.
there is
a

very ductile behaviour. The

plastic hinge region chosen for the development of a suitable

need for tools for the

mechanism has to be detailed for the

Complementary

to the

design procedures
way that

analysis

of

the nonlinear behaviour of structural walls in the time domain. The

must

plastic hinge region

should be extracted

be modelled in such

realistic Simulation of the hysteretic behaviour is

demand in the

possible. Furthermore,
of the

the local

ductility

plastic hinge

for the purpose of comparison with the

design ductility

demand assumed at the

beginning

design process.
doctoral thesis, two kinds of model for in the time domain
are

In this report, which is based

on a

simulating
A

the nonlinear behaviour of

macro

plastic hinge regions

macro

developed:

model and

micro model. The

model consists of four nonlinear

a

Springs
overall

connected
In

by rigid beams.
the
macro

The micro model is

smeared crack finite element

approach.

particular,

model

proved to be capable of realistically simulating the

others the

earthquake behaviour of structural walls.

The results of the

analysis including among

ductility demand,
using

the moment

resistance demand, and the shear resistance demand, may be

directly obtained only

in the

and

compared
of time

to the

assumptions of the design developed

process. Hence,

the modern Computer

models and tools

in this report remarkable progress not

technique
practical

history analysis of reinforced

concrete wall structures but also for

earthquake design purposes has

Zrich, July 1993

been made.

Prof.

Hugo

Bachmann

ACKNOWLEDGEMENTS

This report

was

prepared

at

the Institute of Structural

Engineering

of the Swiss Federal

Institute of Technology

(ETH) in Zrich. Many competent

in this report.

persons contributed conside

rably

to the work

presented

Among them,

some

individuals

ought

to

be

particularly mentioned here.

The author is first of all indebted to Professor Dr. research

Hugo Bachmann, who

as an

made the adviser.

presented

in this report

possible,

and who served

inspiring

Indebtedness is also owed to Professor Dr. Edoardo

criticism. Gratitude is furthermore

Anderheggen for

constructive

especially owed
generously

to Mr. Thomas

Wenk, who served

patiently as

discussion partner and

kindly shared from his experience in the field of non

assisted with the Computer

linear behaviour of structures, and Dr. E.G. Prater's

applications.

proof reading of the manuscript is gratefully acknowledged.

Gratefulness is also
matische

expressed

to

the foundation

Stiftung fr wissenschaftliche, syste

and to the Swiss

Forschungen auf dem Gebiet des

of Technology (ETH)

Beton- und Eisenbetonbaus des Vereins

Schweizerischer Zement-, Kalk-, und

Federal Institute
this report.

Gipsfabrikanten (VSZKGF)

for the generous support of the work

presented in

ABSTRACT

Reinforced concrete structural walls constitute

an

important

unit for the resistance of

structures

buildings against
quakes,

seismic action. In order to

successfully design

against earth typical

to

it is therefore of interest to

develop

numerical model which simulates the

behaviour of these units. This report is concerned with numerical models intended
used in

be

analysis

of

complete buildings,

with focus

on

capacity designed

multi storey

buildings.
A

major part of the report is devoted to highly nonlinear behaviour

the

the

development of a new
of a

macro

model which

simulates the

of structural walls based upon

macro

relatively simple

kinematics and
As
a

physical behaviour. The formulation

to
macro

element is

presented.

complement
treat

model,

micro model is derived with which it is

attempted to
a

the behaviour of the different material components of a structural wall in

relatively

detailed manner, yet also based upon

are

physical observations.
code and extensive tests
are

The models

implemented into

general finite element

presented including comparisons

An

with

experimental data. capacity design

walls of structural walls.
means

important part

of the report deals with the carried out

on

Performance checks

are

capacity designed
demand in the

by

of the

newly

developed macro model.

It is shown that the than

dynamic

curvature

plastic hinge

may be different

suggested

in the

existing capacity design procedure,

take

when varied

over

different wall

aspect ratios. It is further shown that during nonlinear time history analysis flexural

yielding

may

frequently

place in

the upper storeys of the wall which

are

are

intended

to

remain elastic, when the

existing capacity design procedures

used. It is also shown

that the

dynamic

shear forces may be

larger than anticipated by existing capacity design strength

the

assumptions.
An

improved

distribution of flexural

over

height

of the wall is

proposed,

which

clearly reduces

the risk of unintended

yielding

in the upper storeys.

Keywords: capacity design; ductility; dynamic

structures:

structural

analysis: earthquake-resistant
reinforced concrete: shear

finite elements; flexural

strength; hysteresis;
structural

strength; Standards; stiffness; strength;

analysis;

structural

design; walls

TABLE OF CONTENTS

PREFACE

ACKNOWLEDGEMENTS
ABSTRACT

I JJ

1. INTRODUCTION General 1.1 1.2 Objectives and Limitations 1.3 Scope of Report

1
1

2. REVIEW OF NUMERICAL MODELS 2.1 Introduction

2.2

7 7 7

Macro Models

2.3 2.4

2.4

2.2.1 Beam element models 2.2.1 Truss element models 2.2.3 Multiple spring element models Micro Models Meso Models Choice of Models for Development

8
10

10
12

13 14

3. MACRO MODEL 3.1 Introduction 3.2 Model Configuration 3.3 Elastic Hexural Behaviour 3.4 Elastic Behaviour under Normal Force 3.5 Nonlinear Flexural Behaviour 3.6 Shear Behaviour 3.7 Element Formulation 3.8 Ductility Demand

17

17 18
21

25 26
44

51

63

4. MICRO MODEL 4.1 Introduction

4.2

65 65

4.3 4.4 4.5

Material Behaviour of Concrete Aggregate Interlock Material Behaviour of Reinforcement Steel

67
77

80 83

Interaction between Concrete and Steel

5. NUMERICAL EXAMPLES 5.1 Implementation 5.2 Analysis Procedure 5.2.1 General 5.2.2 Time integration 5.2.3 Residual forces

89

89 90 90 90 92

rv

5.2.4 5.3

Damping

93

5.4

5.2.4 Ground motion Selectionof Numerical Examples 5.3.1 Test specimen 5.3.2 Capacity designed multi-storey wall Macro Model Results
5.4.1

building

Comparison with experimental results

Monotonic behaviour Cyclic behaviour

5.4.1.1 5.4.1.2 5.4.2 5.5 5.5.1

Multi-storey wall
to

93 94 94 97 101 101 102 108 115

134

Micro Model Results

5.6

Comparison experimental Multi-storey wall Comparison between Macro Model and Micro Model
results 5.5.2

134 143
144

6. CAPACITY DESIGN CONSIDERATIONS

6.1 6.2 6.3 6.4 6.5

Introduction Flexural Overstrength Local and Global Ductility Demand

Energy Dissipation Flexural Strength 6.5.1 Implication of numerical results 6.5.2 Suggested flexural strength design 6.5.3 Numerical example
Shear Behaviour

6.6

149 149 150 155 160 163 163 167 170 177

SUMMARY, CONCLUSIONS, Summary

Conclusions

AND RECOMMENDATIONS

Recommendations for Future Research

183 183 184 186

ZUSAMMENFASSUNG, SCHLUSSFOLGERUNGEN
AUSBLICK

UND

187
187 188

Zusammenfassung

Schlussfolgerungen
Ausblick

190

REFERENCES

191

NOTATION

199

APPENDDCA

APPENDIX B APPENDIX C APPENDDC D APPENDIX E

APPENDIX F

USER ELEMENT INPUT DESCRIPTION USER MATERIAL INPUT DESCRIPTION YIELD MOMENT FOR MACRO MODEL ELEMENTS FOR MICRO MODEL FREQUENCY STUDY OF DAMAGED STATES DESIGN DEFINITIONS

205
211

217 221 225 231

CHAPTER ONE

INTRODUCTION

1. General

In many

buildings,

reinforced concrete structural walls

such
as

provide

an

important part

of the

resistance

against lateral actions,

wind and

earthquake. Li multi-storey buildings,

means to as

R/C structural walls may often be the

only possible large degree

achieve suffieient lateral

resistance. Tail structural walls act to

cantilever beams, and the lateral

resistance they offer is

mainly of a flexural nature.

The term "shear walls", which is also the

commonly used,

may therefore be somewhat

misleading in that it gives

impression

of

major shear behaviour.

and will hence be used

The term "structural wall" does not lead to such

misinterpretations

throughout this report.

design,
action lead to the failure of

Should, despite
structural wall,
a

careful structural

a severe

flexural failure is

strongly desirable.
and

This is due to the fact that flexural

a

failures

occur

less

suddenly than shear failures,

as

provide

better

means

of

avoiding a
in

structural
and

collapse,

well

as

better

rebuilding possibilities. Furthermore, by consistent

a

carefully performed structural detailing,

structural wall may streich

flexuralry

controllable way far into the nonlinear rnge without

failing.

In the

early days
as

of reinforced concrete structural walls

were

mainly designed

and

analysed

"wide" columns.

During

the 1950's

increasing

interest in the behaviour of

on

reinforced concrete structural walls

the shear behaviour

developed. Experimental studies focused mainly

[T055]

or on

the axial load

carrying behaviour [Lars59].

subjected
to

Increased interest in the behaviour of tail structural walls

lateral action

emerged

in the 1960's in connection with

more

widely spread seismic design provisions,

was

mainly developed

in California. The

design

essentially

still dominated

by

the

assumption of

"elastic" behaviour. The finite element method,

emerging

at

this time,

made Computer simulations of the behaviour of tail structural walls

increasingly feasible,
[Nils68].

although they were essentially

linear Computer

limited to linear programs and to research purposes. Non

analysis

of reinforced concrete structures had just started

CHAPTER ONE

In the 1970's the inclusion of the nonlinear behaviour of reinforced concrete in

design
on

gained

increased international interest. As

continuation of the extensive

in

findings

reinforced concrete
area

published by
Method"

Park and

Paulay

[PP75], substantial advances in the

were

of seismic

design

of reinforced concrete structures

achieved

by

the introduction

of the

"Capacity Design
et

developed

in New Zealand

during the last two decades

are

by Paulay

al,

see

[PBM90] and [PP92]. Further summaries

given by

in

[BP90] and

[MP90]. This method deals with the design of reinforced

concrete structures so as to
a

achieve controllable ductile behaviour. The method has been confirmed

of experiments, Zealand. The

large number

mainly carried

out at the

University of Canterbury by
selection of

in

Christchurch, New

preferable
The

ductile behaviour is achieved

plastic hinge

zones,

careful structural

detailing

of these, and
zone

protection

of the rest of the structure

against

yielding.

plastic hinge

of structural walls is

usually

located at the base of the

wall, while the

rest of the wall is intended to act

essentially elastically, although it may

with dimensions

become cracked.

Figure

1.1 shows

typical multi-storey capacity designed building height H, in

the

length L, depth D, plastic hinge

the

and

which structural walls resist horizontal actions. The

over

of each wall is located at the base and Stretches

length Lp, taken

as

largest of H/6 or the wall length Ly,.

Structural wall

aaata

H
Sgs

--

im
Plan
L-_<

Section

Figure

1.1

Typical multi-storey building with structural walls for the

resistance

of

horizontal actions

INTRODUCTION

Confinement hoops,

tighy spaced in plastic hinge

zone

Ly,

a) Cross section with protruding boundary elements

Z.

Flexural reinforcement bars

b) Rectangular cross section

Figure
Figure

1.2

Typical wall cross sections for capacity designed walls typical

sections of
are

1.2 shows two

cross

capacity designed multi-storey

are

structural walls. Flexural reinforcement bars

concentrated at the ends and

zone.

confined

by hoops which

are

tighy spaced

in the

plastic hinge

Sometimes

protruding

boundary elements may be necessary in

with
as

order to resist

large normal forces in combination usually placed

recommen

bending

moments.
as

Vertical reinforcement in the "web" of the wall is

to

economically

possible, often according

minimum rules. Detailed

are

dations for flexural-, shear- and

confining-reinforcement placement
process, where materials
are

provided

in

[PBM90].
In the modern structural

design

are

used

efficiently

and for

which the demands for

the

safe and reliable end result

increasing, numerical tools for

in the verification
a

analysis of the structural behaviour constitute a very important aid

of the

Performance

of the

designed

structure.

For

complex multi-storey buildings

the above mentioned the nonlinear

rational

design is usually no longer possible without the aid of Computer programs. Also, design methods,
such
to
as

when consistent and rational

Capacity

Design Method,
means

are

used, it is of great value

verify

Performance by

of a Computer program.
as

For R/C structural walls, however, there is

and

of yet, among the extensive commercial any

university Software and despite numerous promising attempts, hardly

numerical model

simple

and

readily available

designed specifically

to

simulate the nonlinear

behaviour of structural walls under seismic action.

CHAPTER ONE

1.2

Objectives

and Limitations

The

smdy presented in this report should be

seen as a

further attempt to

clarify

the

most

important features of the nonlinear behaviour of structural walls by

In order to

means

of numerical

modelling. Emphasis is placed on Performance control of capacity designed walls. accomplish this, the
the

work must be

organised

so as to

fulfil

number of

objectives
set out as

set out at

beginning of the study.

The

major objectives of this report were

follows:

1. Based upon

an

overview of the State of the

a

art

of

major

numerical wall

models to select

model

worthy of further development, intended for

and regulr wall geometry. structural

simulating multi-storey wall buildings with simple

This model should in its final State be

capable of analysing various

properties relevant to the capacity design method.

2. For the control of the behaviour of this model well
as a

during dynamic analyses

of involved and

as

complement
a

to

this model in

cases

irregulr

geometry etc.,

second model should be

a

developed, describing the material

manner.

behaviour of the constituents in model should

not

relatively detailed

The second

in the first

place be designed for control of capacity design

Performance.
3. To

perform

detailed and transparent

development

of these models into

workable numerical tools for nonlinear structural wall

should be

analysis. Emphasis
when
no

placed

on

simple solutions,

for both models,

particularly

proof of the superiority of complex solutions exists.

4. To check the

reliability of the developed models important and which

in

number of numerical and

examples including comparisons with experimental data,

model parameters
are are not

identify which
to compare

Further,

the

responses of the two models for

dynamic behaviour of multi-storey buildings.

of

5. To review the structural

means

Performance

capacity designed buildings by

to

of the model

developed for this purpose, and consider it in relation

basic criteria

speeified at the design phase.

INTRODUCTION

6. If applicable, to suggest

some

general improvements

in the

capacity design

procedure

for structural walls.

7.

Finally,

to

give recommendations for further research.

In order to

perform this smdy in

an

efficient

manner a

number of limitations have to be

imposed on
follows:

the extent of the work. The

major limitations

of this report

were set

out as

1.

Mainly

the

global behaviour

of structural walls is of interest.

In-depth

analysis of local phenomena, other than the degree

development, is not carried out
2.

necessary for the model

Except for gravity loading,

as

this report focuses

on

earthquake action only.

Other actions such

wind, creep, shrinkage, temperature, environmental

effects, etc.,

are not

considered in this

smdy.
the nonlinear behaviour of walls
no

3. The

modelling

in this report focuses

on

from the outset. Therefore, there will be

essentially

discussion of linear

dynamic analysis results.

4. The does

smdy

limits itself

to

the behaviour of structural walls and

essentially
to

not

deal with

problems associated with the

as

connection of walls

other

structural elements such

beams and columns.

5. The work does not enter into the

problems of major irregularities of walls:

well the
as

openings in walls, coupled walls,

such
as

as

three-dimensional
of

configurations
such
cases

lift shafts, etc.,

although

capability

modelling

should be

regarded in

an

overall evaluation of a model.

6. Interaction between superstructure and soil is

beyond

the scope of this

on

study, although Performance

in

it is known to have
some

an

important influence

the structural

situations, e.g. soft soil.

CHAPTER ONE

1.3

Scope of Report Organisation

of the report is

The

as

follows: the second

After this

introductory first chapter,

models divided into and

chapter is

devoted to

review of

different numerical models for the Simulation of nonlinear behaviour of structural walls.
The

existing

are

major categories, disadvantages

with
a

and for each category

are

brief

background

is

given. Advantages

of each model category

briefly

discussed. The second

to

chapter concludes

selection of two numerical model types

be further developed in this report

Chapter

three deals with the

development

of

numerical model intended for the

Simulation of multi-storey
as a macro

capacity designed walls. This model is classified in chapter two

macro

model. An efficient and transparent

model is

developed

in stages,
as

including

the derivation of kinematic relations, flexural and shear behaviour,

as

well

as

axial load effect. Cross sectional characteristics such the model.

A micro model is

ductility ratio are implemented in important

is

developed in chapter four.

in
a

After identification of the most

phenomena
the model.

to

be

regarded

micro model, the crack behaviour of the

as

concrete

developed. Then the

Numerical
include

steel behaviour

well

as

the concrete-steel interaction is included in

testing of the developed models is presented

with

in

chapter
on

five. These tests

comparison

experimental data,
In

as

well

as a

check
some

the

Performance

of

capacity designed multi-storey building.

chapter six
are

aspects of the capacity

design

method in relation to numerical results

seven,

discussed. The report concludes with

chapter
The
as

containing

summary, conclusions and recommendations for future

research.

developed numerical models

were

implemented into
In the

general finite element code

of this report there

"user elements" and "user material",

a

respectively.
use

Appendix

is

short users' manual,

describing the

of these models.

CHAPTER TWO

REVIEW OF NUMERICAL MODELS

2.1

Introduction
many different models for the numerical
a

There

are

modelling

of structural walls found in

some

the literature. As

background to
advantages

the

subsequent chapters

of this report

of the

more

important numerical models are presented in this chapter. The different models are briefly
described and their
and limitations mentioned. For
a more

comprehensive
The of a

study

of any

particular

model the reader is referred to the

seen

appropriate reference.
to the

discussion of the different models is here

mainly relation
as

analysis

global

building statically

structure,
or

subjected

to

gravity loading

well

as

earthquake forces applied

dynamically. modelling
of structural walls may, from

The numerical

structural

engineering point
section

of view, be classified in two

major model

levels:

macro a

models and micro models. The

cross
over a

former attempt to model the overall behaviour of

structural wall

certain

height,

while the latter base the behaviour upon the constitutive laws of the

mechanics of solids. Models which may be be referred to

as meso

placed between the two major categories

on

may is

models. A

good discussion
[VB87].

the two

major model categories

provided by

Vulcano and Bertero in

It may be

especially

noted that the

definitions used here should not be confused with similar definitions used in the field of

fracture mechanics.

After

description

of the different numerical models and their

applicability to

various

structural

analysis situations
development in

this

chapter concludes

with

discussion of a suitable model

for further

this report.

2.2 Macro Models

The category "Macro Model" is here understood

as a

numerical model which attempts to

a

incorporate
height
or

the entire behaviour of

major region

of

structural wall, such

as

as a

storey

part thereof, including the wall's constiments such

the concrete, the

reinforcing steel,

and the interaction effects between concrete and steel.

CHAPTER TWO

'/////////////ZA

Y///Y/////A
O Node with translational
and rotational d.o.f.

Jk~^
Figure 2.1

Structural wall prototype

Figure 2.2

Beam element model

In the Hterature several different

macro

models

are

found for structural walls. However, it

appears

possible

to divide the more

important

and

frequently

used models into three

types, which will be discussed separately in the

following.

2.2.1 Beam element models

The

simplest numerical

model for

structural wall consists of beam elements, with six

case

degrees of freedom
is
a

per element The wall in this

and in
some

is

regarded

as a

deep column.
a

This

very

commonly used concept sufficiently unimportant

analysis situations it may provide edges

model

which is
are

realistic. If the vertical deformations at the wall

or are

due to flexure
one

considered

assumed to be small, the entire wall model for element.

storey may consist of a

single beam
seen
a

For the prototype wall

in

figure

2.1 the beam element

modelling
as

is shown in
case

figure 2.2.

For walls with

a

considerable horizontal

length,

as

well

in the

of

an

interaction with

structural frame, it may be necessary to consider the vertical

edge

deformations. A

simple Solution including this effect has been suggested by adding hori-

REVIEW OF NUMERICAL MODELS

a1-

Stiff Beams
h

r-

--

It

9//////////TA

-w

O Node with translational

and rotational d.o.f.

Node with translational d.o.f.

Figure

2.3 Beam element model with

Figure

2.4 Truss element model

horizontal stiff beams

zontal

rigid beams on either side

of the vertical beam [BASC84],

as

shown in

figure 2.3,
and
Few

and

thereby obtain vertical deformations at the wall edges.

advantages of these
beam models consist of the

The

uncomplicated modelling,

sometimes

possibilities

to

check the results

by frame analogy hand calculations.

in

degrees of freedom is another advantage, especially

The limitations
cross

dynamic analysis.
along
its

are

mainly

due to the

inability to
at

describe the wall's behaviour of the wall

are not

section. The vertical deformations

are no

each

edge

considered if

there

horizontal

rigid beams.
and

Even with these

rigid

beams the strain distribution

will

not

be

realistically modelled, cracking

since the shift of the neutral axis, which is

oceurs, cannot be

typical

for

wall when flexural is

subsequent yielding

reproduced. This

especially noticeable
considered

under flexure at the tensile

edge where the large tensile strains are

not

by

the model.

10

CHAPTER TWO

2.2.2 Truss element models

The next

macro

model is the truss model, of which different versions model

as

are

presented.
elements,

Typically,

a truss

shown in

figure 2.4,
are

consists of connected

two
a

vertical

truss

and at least one

diagonal

truss

element. These

by
al

rigid horizontal beam.

and

Truss models like this have been used e.g.

in

by Vallenas et

[VBP79],

by Hiraishi
strut"

[ACISP84].

The

diagonal

truss

is

supposed to model the concrete "compression

diagonal
in the

which forms under lateral force. This behaviour may be

under force reversal it is necessary to
use a

reproduced quite well, however,

truss

opposite diagonal

direction. Furthermore, the

reproduction of behaviour under

as

various moment/shear of deformations due

to

applications

seems

problematic,

well

as

the realistic

modelling

gravity load and lateral force, each by itself,

For static monotonic force

and combined.

application, and

for

small
its

gravity load,
use

the model may

to

provide
the

useful results, if

carefully calibrated. However,

a

appears

be limited to

rather squat walls, where

compression

strut

of this nature
to

aetually develops. Further, dynamic

versatility

of the model may be limited

appear feasible.

compared

other models, and

analysis does not

2.2.3

Multiple spring element models

macro

The third

model is the

multiple spring element model, which originated in the early

US-Japan Cooperative Earthquake Research Program
modelling in plan
the

1980's within the framework of the

[ACISP84]. The model

the

was

intended for the wall

analytic prediction of
out at

experimental

tests on a

full-scale

seven

storey R/C structure, carried

of the
test

the
see

Building Research
figures

Institute in Tsukuba. For elevation and

specimen,

2.5a and 2.5b.

The first numerical model of this type, the

suggested by Kabeyasawa et al,

as seen

was

used for

modelling of single storey wall

all connected

sections
one

in

figure 2.5c, spring,

and

and is shown in
one

figure

2.6. It

comprises three vertical Springs,

are

rotational

horizontal
seven

spring,
test

which

by rigid beams.

The nonlinear behaviour of the

storey

structure could

be simulated quite well.

characteristics of nonlinear structural wall behaviour, such
as
as

Generally, important
large
can

tensile strains,

shiftmg of the neutral axis,

based

well

as

signifieant

shear deformations,

be simulated

adequaxely by models

on

this

approach.

REVIEW OF NUMERICAL MODELS

11

I
s

i
i

I
i

VO

-ir*
VO
II"

8"
CS

(.
6.00

1.
5.00

J
6.00

b)Plan
I

yYv////Y^YfYyYr//////y/

----4
I

6.00

$1
I
I

5.00

6.00

\
test structure

I
I

a) Elevation

c) Macro Model

Figure 2.5 Seven-storey reinforced concrete

with structural

wall, tested in

fll scale at Tsukuba [PCASP84]

m VMMMMM^^^^^

Vsssss/wss^^^

&
model

Figure 2.6

Initial

macro

simulating one storey ofthe structural wall

ofthe seven storey reinforced concrete test structure at Tsukuba [PCASP84]

Some

major

limitations of the model

are:

the

rigid beams imply that plane

of which is the main

cross

sections remain

plane which is

poor

assumption for deep beams and walls, but less

smdy

critical for very tail and slender walls, the

objective
and

here.
even

Experimentally obtained strain distributions of a slender wall specimen [WK85]

12

CHAPTER TWO

of

relatively squat specimen [VBP79]

to

show that in

mainly

flexural modes the

cross

sections remain close

is
not

plane

even

far into the nonlinear rnge. Furthermore, the model

capable of taking into

account a

bending moment gradient along its element height,

on

and it does not

provide mueh information

localised

damage, such

as

crack direction.

Nevertheless, the model appeared to give reasonable agreement with

data.

some

experimental

Refinements of this
with
more

original model

have been

attempted by

some

authors, and is dealt

fully in chapter three.

2.3 Micro Models

The category of Micro Models is based upon the mechanics of solids, and wide field of the

comprises the
in

modelling

of constitutive relations, and their

implementation

continuum elements. In the

may be

case

of structural walls, with the usual

the

approximations,

this

performed by applying

plane
as

stress

relations of the materials and

by

implementation in
approaches
smeared for

membrane elements,

shown in

figure 2.7

During the early research of nonlinear concrete behaviour in the late 1960's two major

modelling cracking

of concrete evolved: the discrete

approach

and the

approach.
to

It has been found that the smeared crack

approach lent itself more

with its interaction effects

efficiently

modelling the behaviour of reinforced concrete

between reinforcement and concrete and well distributed cracks of moderate crack width. On the other hand, the discrete crack and Nilson
concrete

approach, pioneered by Ngo

to

and Scordelis
structures

[NS67]
such
as

[Nils68]

was
a

found

to

be well-suited

unreinforced
a

dams, where

few cracks with wide

openings play
the These

signifieant

role in the

changed

structural behaviour. For the discrete

approach

problem

of mesh

updaxing
still hold

has been treated among others

by Skrikerud [Skri82].
on

findings essentially

today, although

the discussion

this

topic continues.
introduced

The smeared crack

concrete

approach

was

by Rashid [Rash68] for the analysis of

pressure vessels. The first attempts at nonlinear

analysis

of structural walls
An

by

the smeared crack

to

approach date back to around 1970, [Cerv70, Fran70]. by

their

application given by

the

global analysis of tail structural walls by membrane elements

[ACISP84] for

was

Moazzami and Bertero in 1987 [MB87],

the concrete test structure at Tsukuba
der and

modelling of the seven-storey wall of

monotonic load conditions. Bolan-

Wight [BW91] analysed

ten-storey

concrete

building

with several structural

walls under monotonic

loading. During the

seventies

andeighties, further efforts went

REVIEW OF NUMERICAL MODELS

13

i
/

a) Discrete approach

b) Smeared approach

Figure 2.7

Micro

modelling of cracking

zone

in

wallpanel byfinite elements; different

crack model approaches

into

modelling

the behaviour of cracked concrete and the interaction between steel and
to

concrete.

However, relatively few attempts have been made

develop
to

simple

micro

model which exhibits reasonable

global behaviour under seismic action.

are

Clear
on

advantages of the micro model

more

its

versatility and ability

give information

localised behaviour. The

elaborate model

generation and higher numerical effort

compared to macro models however, is a clear drawback. For multi-storey buildings with
several structural walls
a

comprehensive dynamic Simulation may not be feasible. Another

by design engineers
as

limitation may be the lack of interest shown formulations often

in the rather involved

presented
use

in many reports

well

as

the lack of

global results,

resulting in
described

limited

in

practice.
element codes

Some of the
as

commercially-available finite
be that this is
in detailed

provide

material behaviour
A

e.g. "concrete behaviour" to be used with

existing library elements.

problem

seems to

usually only intended for monotonic loading, and thus

earthquake analysis comprising cyclic
to
or

may not be of

use

dynamic
user

behaviour. Furthermore, the freedom

modify

this

concrete

behaviour

by

the

is

usually limited.

2.4 Meso Models

A category of models which may be models is

placed between

the

macro

models and the micro

as meso

presented by

e.g. Meskouris et al
as a

[MKHH91], and
to

will be denoted

models. In this reference

complement

detailed micro models,

simplified

two

14

CHAPTER TWO

dimensional wall models

are

presented.

The

justification for
to

these models

was

mainly
bilinear

given with regard

The
meso

to

computational efficiency in comparison

the

more

detailed models.

models consist of two dimensional membrane elements, with

simple

material models. material

Explicit

crack formulation is not taken into account

by orthotropic

expressions,
and

but instead

simplified hysteretic rules

implemented
more to cases
a

are

used to account for

cracking
Thus,

yielding.
models, although they
are

meso

in continuous elements,
macro

use

simplified
slightly
and

material behaviour which

belongs
In

the

models. The results of

these models may be of

varying quality.

where the overall behaviour is

only

nonlinear these models may represent effort.

good compromise between Performance

computational

2.5 Choice of Models for

Development

The numerical models described above all offer particular

advantages

and

disadvantages

for

given analysis
The

tasks. It appears difficult to find

to

model which displays

some

only or mainly

advantages.

opportunity

compare the results of

different numerical models

applied to the

same

problem should be helpful for complex analyses.

In this report the

analysis of the global behaviour of tail slender structural walls under primary interest. Thus,
they
function
to a

seismic excitation is of

the models

on

the macro-level appear

a

suitable for this purpose, since

large degree in
the

global manner.

Of the

three types of

macro

models described

previously,
of

multiple spring model generally

the main characteristics of

appears to be the most

promising

and

capable
well

simulating

nonlinear wall behaviour.

Its

predominant global behaviour,

the

as

as

relatively easily defined

cross

sectional

Output

quantities, make it suitable for the analysis of capacity designed walls. multiple spring model
is referred
to as
was

Consequeny,
the macro-level

selected for further

investigation

in this
on

report, and from

now on

the "macro model". The other model types

are

treated in this report

The

reliability of the selected macro model may

data. However, since the nonlinear

be checked for static behaviour response of

a

against
wall

experimental
to
a

dynamic

multi-storey

buildings cannot be reproduced experimentally reliably (on

comparison basis for dynamic problems.

large scale), it is necessary

complement the multiple spring model by a clifferent numerical model, in order to have

REVIEW OF NUMERICAL MODELS

15

Macro

bmparison of
dynamic behaviou

model

Performance tests

>

Capacity designed
structure

(Chapter 3)

(Chapter 5)

Experimental data

from static tests

Capacity design procedure

(Chapter 6)

(Chapter 5)

Figure
Furthermore, for

2.8 General view

of models and analysis objectives

some cases

of

global wall analysis, mainly

seems
a

with

irregularities
analysis

of of

various kinds, the micro-level model

a

preferable.

When

a more

detailed

particular region

of a wall is of interest,
an

micro model also

seems to

be favourable. In

order to fulfil these purposes

attempt

to

develop

relatively simple

and transparent

micro model for structural walls will be within the framework of this report. This micro model should be based
suitable for the
on

the smeared crack

approach,

which appears to be the most such

as

modelling

of

uniformly reinforced

concrete structures

structural

walls.

Although clearly

the

meso

models may be useful in

some

cases,

especially as a complement to
non

micro models in the

more

same

analysis,

e.g. for

regions which behave less

seem to cover
meso

linearly,

the

defined

macro

and micro models

the essential

points

of view

in the discussion of wall models.

Therefore,

models will not be treated in this

report.
The

development of a macro model is presented in chapter three, and the development

chapter four. Separate reliability tests
performed in chapter five.
wall is
as

of a micro model in

well

as

comparison of their

respective dynamic
of
a

behaviour is

The actual

Performance check
use

multi-storey capacity designed

on

performed mainly by

the

of the

macro

model, and is partly based

results in

chapter five,

but discussed in

more

detail in

chapter six.
Based upon
some

findings

from the nonlinear time

history analyses using

the

macro

model,
also

modifications in the
in

capacity design procedure

in

for structural walls

are

presented placed on

chapter six.

general

view of the selected models and

analysis objectives
will be
well
as

with these models is

the
area

presented

graphical form
testing

in

figure 2.8. Mueh emphasis

designed
wall structures,
as

of Performance

of capacity

the discussion

on

improvement ofthe design procedure.

16

17

CHAPTER THREE

MACRO MODEL

3.1

Introduction

This

chapter is devoted to the development of a numerical model for the

multi-storey
structural walls

Simulation of the
seismic action.
macro

overall nonlinear behaviour of

subjected

to

The model works with nonlinear

Springs

and

belongs

to

the category of
macro

models.

Of the three types of model

belonging
nonlinear

to

the category of

models discussed in
beams
a was

chapter

two

the type based

on

Springs connected by rigid

tool for in the

found

to

be the most suitable for the Simulation of structural walls. In this

chapter,

model of this

type is developed into

this report.

functioning numerical

use

subsequent chapters of

A model of this type

was

was

originally suggested by Kabeyasawa et

was

al

[KSOA82], and

shown in

figure

2.6. This model type

used

during

the

US-Japan cooperative
to

earthquake research
a

program

during

the 1980's, and its of

primary objective was

structural walls

provide
to

simple

tool for the nonlinear

analysis

multi-storey
and

subjected

earthquake
sections.

actions. Some further

developments
recently

of this model type have been made

are

by

other authors in the late 1980's until

briefly

discussed in the

following

In this

chapter

the

geometrie considerations

are

discussed first based upon

a

short

review of

previous

work

by

different authors, and

on

simplified

and efficient model

geometry is suggested. Based

behaviour
are

kinematic relations the basic model

on

properties for elastic

derived, and based

the well known material behaviour of concrete and

steel, combined with observations of physical behaviour, the nonlinear flexural properties
are

developed.
are

Axial and shear behaviour

are

treated

separately. Simple

and efficient

hysteretic

rules

developed largely

based upon

empirical observations. Finally, the formulation

presented.

ofthe

stiffness matrix of a

macro

element is

18

CHAPTER THREE

3.2 Model

Configuration
model

The

original

by Kabeyasawa

et

al, shown again in figure 3.1, consisted of five

nonlinear

Springs, connected by rigid

Springs, representing

beams. The

Springs

were

made up

as

follows:

two
one

vertical outer

the axial behaviour of the

boundary columns,
one

central vertical

spring, representing

the axial behaviour ofthe web,

central horizontal
one

spring representing
rotational

the shear behaviour of the wall section, and

finally

central

spring,

intended to represent the flexural behaviour of the web. The three

or near

central

Springs were located at the base of the element,

of the
was

the base.

Each

one

seven

storeys of the fll scale

test

specimen
an

in Tsukuba,
as a

presented in

chapter two,
as

modelled

by

a set

of Springs to form

element used

storey model

shown in

figure 3.1.

In later

developments
was

models based in this type have been used in

more

examples
aecuracy,

in which each storey

see

discretised into

than

one

element for better

e.g.

[BWL92a].

VtfSSmWjWjWjM/mW^^^

svWMMMMMWMm)m/Mmmi/mM//nuM &
Figure
The
the

3.1

Original macro model by Kabeyasawa et al [KSOA82]

al

original model by Kabeyasawa et

of the static

was

essentially used in

numerical

analyses for

prediction

cyclic

and

pseudo dynamic

tests of the fll scale wall in

Tsukuba, and the scale models in the US-Japan cooperative research program.

The attempt to

separately model flexural

and axial behaviour in this

manner

led to

compatibility problems, discussed by Vulcano and Bertero [VB87] and Linde [Lind88],
[Lind89]. These difficulties arise mainly when flexural and axial properties
to
are

assigned

the rotational and vertical

Springs, respectively,
base
on

as

suggested for the original model,

web and the

since these

assigned properties

the

independent behaviour of the

MACRO MODEL

19

boundary columns. Attempts

softening
fully
stiffness
or

were

made in

[VB87]

to correct this

problem by assigning
this Solution
was

properties

to

the rotational
seems

spring. However,

not

reliable

efficient. It also

difficult to

explain

the

softening stiffness

physically.
Furthermore, it

was

attempted

in

[VB87]

to

model the outer vertical and

in

Springs by

spring assembly
was

which simulates the

a

physical behaviour of cracking spring assembly,

concrete
as seen

yielding.

This
The

represented by
on

parallel

and in-series

figure 3.2a.

single spring

top is intended for uncracked

and steel,

while the

parallel Springs
a

below

model cracked

concrete

respectively.

The steel

spring follows
or

bilinear curve,

and the concrete crack-spring either seizes to act of crack). This

(cracked State)

takes up action

(closing

interesting approach may be seen as an attempt to kinematically model the

material of

physical behaviour rather than employing hysteretic models for the composite
reinforced
concrete.

^^^^^^^^^^^^^^^^jj^.y^^^^^A

mm
WMM/SM. >jS,}s>;//sa zzzzzz

a) Parallel & in-series model [VB87]

b) Multiple vertical spring model

for vertical outer Springs Figure

3.2

[VBC88Jand[FF91]

Suggestions for improved macro models

Fajfar
and

Vulcano et al [VBC88] and

Fischinger [FF91]
to

then

replaced

the rotational

spring by

several additional vertical

was

Springs

simulate the axial behaviour of the web.

This method

able

to

simulate the
more

gradual yielding of the vertical reinforcement more

to a more

smoothly, but it consists of

model.

components and thus leads

complicated

Generally,

refinements lead in the direction of micro models.

20

CHAPTER THREE

Attempts have also been made

the
macro

to

develop simple

and clear kinematic formulations of

was

model. In

[Lind88], [Lind89] simpler geometry

suggested

and tested for

static

loading.

In this

stady this concept will be continued geometrie spring arrangement

to
as

and

model will be

developed,

which is

based

on a

shown in

figure 3.3.

The idea behind this

a

arrangement is

omit the central rotational

spring

and to

perform

derivation of the
axial and flexural

properties

for the

remaining

three vertical

Springs

so as to

satisfy both

behaviour. The horizontal

spring, modelling

shear behaviour, continues its function,

as seen

making

total of four nonlinear

Springs connected by rigid beams,

in

figure

3.3.

The model thus fulfils the necessary and suffieient

spring arrangement

in order to

simulate the most important kinematic wall behaviour.

The idea behind the arrangement in few

figure

3.3 is thus to achieve

simplieity by using as

Springs

as

possible.
next

The flexural behaviour which is treated in detail for elastic

behaviour in the

section, canbe madeto simulate typical wall flexural behaviour

y^/////////////////mV///j/jj///^

bMMMWMWWMWM>m>/JM/MWM/Vm/f/.

Figure

3.3

Suggested simplified macro model,

only
the two outer vertical the beams
are

based on model in [Lind89J

quite accurately
central vertical

with

Springs,

in combination with the third

spring. Since

connecting

the nonlinear
same as

Springs

are

flexurally
more

rigid,

the kinematic

possibilities

essentially the

for models with

complicated spring configurations, and thereby

this model is able to

provide

an

efficient

result with the nonlinear behaviour derived properly.

MACRO MODEL

21

3.3 Elastic Flexural Behaviour

For the wall model which

was

shown in

figure 3.3,

we

derive here realistic

properties for

the elastic flexural behaviour. Axial behaviour, inelastic flexural behaviour and shear behaviour will all be treated in

subsequent sections of this chapter.

have the two outer vertical

For the flexural behaviour

we

Springs

at our

disposal,

and

let them simulate the flexural behaviour of the entire wall section, i.e. the web of the wall,
and

boundary

columns

(if present), together giving

certain moment of inertia and

area.

We consider first the elastic behaviour of the wall model. This is

preferably figure
for

accomplished by comparing
model. For this purpose,

the kinematic relations of a

simple real wall with those of the

as

we

consider the cantilever structural wall

shown in

3.4a, which

we

refer

to as

the "real wall", and which

obeys simple

elastic

theory

beams. Under pure flexure the model in

curvature.

figure

3.4b would have to simulate the uniform

////////YY///////

V
Figure 3.4
By expressing

1
b) Model

a) Elastic theory

Kinematics

ofwall rotation for uniform moment distribution displacement at the top of the wall for the two
some

the rotation and

we can

walls in

figures. 3.4a

and b,

derive

simple model parameters.

moment

The real wall has the

height h, length /,

cross

sectional

area

A,

of inertia /, and

Young's modulus E.

22

CHAPTER THREE

For the model real wall and

outer
we

we

may then,

by prescribing

the

same

rotation and
to

displacement
area

as

the

assuming
as

the distance between

as

outer

Springs

be /, obtain the

of the

Springs, As,

well

the location for the centre of rotation,

of the model,
so as to

hc.

For this purpose,

lock the horizontal

spring

obtain

only

flexural deformations.

For

figure

3.4a

we

obtain

R-Mh

e-~kT

(3-D

8h=^*

2EI

(3.2)

and for

figure 3.2b, with F.

=,

-,

and

*=

LA

we

get

Mhhc
Sk-ehc=^tl2A.E

(3.4)

By setting (3.1) equal

21

to

(3.3),

and

(3.2)

to

(3.4)

we

obtain the model

properties

K--w l2

<3-5>

K~\
representing
the outer

(3-6)

spring

area

and the relative centre of wall rotation,

respectively.

The elastic stiffness

Ku

of the outer

spring

is then

given by

*=*By replacing
the moment M with
a

<3-7)

shear force V,
as seen

acting

at

the top of the wall, and obtain

a

using the model properties derived above,

in

figure 3.5
gradient.

we

check

on

the

elastic flexural behaviour for shear force with moment

MACRO MODEL

23

/YYYYYYYY////////

\
Figure 3.5

'
a) Elastic theory

b) Model

Kinematics

ofwall rotation for shearforce

with moment gradient

We thus maintain the

properties derived through equations

3.1 to 3.7, and from

figure

3.5

we

obtain for the real wall

by elastic theory

Vh2
=

2EI

(3.8)

Vh3
=

3EI

(3.9)

and for the model

2/

(3.10)

=
v

Fh

V*1*1
=

AEI

(3.11)

e=2A=Yh?_
l 2EI

(3.12)

8h

dhc

}2L
AEI

(3.13)

24

CHAPTER THREE

Comparison
This is due
account

of

equation 3.8

to

3.12 shows agreement for rotation, but the deflection

given by elastic theory (equation 3.9)

to

(liffers from the model's deflection

(equation 3.13).
to

the fact that the model derived for pure flexure is not able

accurately
be reduced

for a

moment

gradient over the height

several elements,
to

This

deficiency may, however,

and

by discretization using
shear force V only,

[Lind88], [Lind89].

Another possibility would be

derive the

quantities A.
in

hc using

the

case

with

acting on top

of the wall,

as

figure 3.5.

The values for rotation and

deflection at the top would remain

For the model
we

according

to

equations 3.8 and 3.9

for elastic

theory.

obtain

F.

^=

(3.14)

*=TF

^!
IA.E

(3'15)

e=2A
l

2VhhL ?A

Sh

Bhe

-^ir
PA

IVhh]

(3-17)

Equating
4Ht 3/2

the

expression

3.8

to

3.16 and 3.9

to

3.17,

we

obtain the model properties

(3-18)

K=2J
This shows that for the
case

(3.19)

of shear force

only

with moment

gradient

the centre of
case

relative rotation would be located lower (hl3 from the base) than for the flexure (h/2 from the base). This
moment
means

of pure
have

that in

general
and the

case area

where

we

always

and shear, the centre of relative rotation

hc,

of the

outer

Springs As

depend on the ratio of moment to shear.

Since in

reality

we

always

have both moment and shear

acting

at a

wall section,
The

possibility would be
would

to

combine the results from the two extreme

cases.

simplest
which

method of combination would be to take the average of the above

two cases,

give

MACRO MODEL

25

A<

[f f]/2 f|
+
=

(3.21)

more

sophisticated

way would be to combine the model

or

properties according
to

to

expected

ratio between moment and shear force,

possibly according

the

instantaneous ratio.

Equations 3.20 and 3.21

then take the following form

A*

al2+b%

4/

l-

3/2

(3.22)

hc

a^+b^
4

(3.23)

a +

(3.24)

However, since the rnge between the

effort does
to

two extreme cases

is

already quite narrow, the modelling considerations going

not

appear justified in relation

to

other more essential

be described.

Since the above

assumptions

are

only valid

in the elastic rnge, and

we are

to

examine structural walls which derivation based

on a moment

are

mainly acting in a flexural State, we concentrate on the

at

acting

the top,

as

shown above. Discretization would

as

reduee the

displacement error for pure

as we

shear action,
more

discussed in [Lind88]. Nonlinear

than the

considerations will,
discussed above.

will see, be

important,

analytical differences

3.4 Elastic Behaviour under Normal Force

The

spring stiffhesses of the

two outer

Springs, derived in section 3.3,

do not account for

the entire axial behaviour alone. A third

spring stiffness,

that of the central vertical

spring,
each

is needed to

accomplish this goal.

a

In

figure 3.6,
cause

the real wall and the model

are

subjected to

normal force N, assumed to

uniformly distributed compressive axial

deformation. For the elastic

theory we obtain

*.~ EAmm,

(3.25)

26

CHAPTER THREE

E, As J

V///////////////
a) Elastic theory

b) Model
Behaviour under normal force

Figure 3.6
and for the model

we

correspondingly obtain
N

N
-

"'

IKst + Kcs

~2EAS

(3.26)
_

where K.. is the elastic

compressive spring stiffness

a

for the central vertical

cross

spring. By

introduction of the ratio

a
=

between

one outer

spring

and the

sectional wall area, i.e.

As I Am,,

the elastic

spring

stiffness for the central vertical

spring may

be written

Ka

EAw

l-2a

(3.27)

Together with

the outer

Springs

this

spring

accounts

for the axial behaviour in

compression and thereby

the

complete elastic

axial behaviour of the model is determined.

3.5 Nonlinear Flexural Behaviour

To determine of the model

properties beyond
cross

the linear elastic

a

region

it is necessary to
we

smdy

the

physical

behaviour of the

section of

real wall. For this purpose

a

establish
cross

a moment-curvature
a

relation, by

means
see

of

Computer program, dividing

3.7. In each fibre, the
are

wall
and

section into

finite number of fibres,

own

figure

concrete

vertical steel 3.9.

obey their

constitutive relations. These

shown in

figures

3.8 and

MACRO MODEL

27

Sk_ Centroid of
cross

section

Fibre

no.

Distance from centroid

to center of fibres

Figure 3.7 Fibre model for cross section

0.012

Ec

Figure 3.8
For the concrete

Concrete material model

we

use

the relation
as

given by

Kent and Park

[KP71] for the

compressive behaviour,

stated

follows

fe< 0.002} -*/.=/; {ec >0.002}->/c

Z
=

0.002

U.002J

(3.28)

/c'[l-Z(ec -0.002)]

(3.29)

0.5

50-50*-0.002

(3.30)

28

CHAPTER THREE

where eSOu and 50A represent the strain the unconfined and confined

at

50%

strength
seen

on

the

descending

branches of

(hooped)

concrete,

in

Figure

3.7. The

following

relations determine f^ and Xk

50u

3+0.002/; /;-iooo

(3.31)

0.75^-4(3.32)

where ps is the confinement ratio,

bh

is the width of the in

hoops, and sh

is the

spacing

of

the

hoops,

and

equation 3.28 assuming fc

psi.

No difference in

strength

is assumed
was

for confined and unconfined concrete,

although

an

increase of perhaps 10%

found

by

some

researchers.

Cracking
steel

was

assumed for each fibre

having

reached the strain

0.0001.
For the

reinforcing

bilinear elastic

(i.e. linearly strain hardening) model

user.

was

used, shown in figure 3.9. The hardening ratio may be chosen by the

In the fibre
a

model, the web steel

which is

area

is smeared out and each fibre of the web obtains

the fibre
area.

steel

area

proportional

to

For the

boundary

elements

a more

refined

modelling is possible
fibre.

so

that

essentially

each

single

bar may be allocated to the proper

fu

'

fy

Experiment

Idealised, used for model

0.05

0.10

0.13

Figure 3.9 Steel material model

MACRO MODEL

29

The Computer program for

cross

sectional curvature behaviour

employs
at

an

incremental
as a

iterative

procedure

described uniform

as

follows.

Any gravity

load is

applied

first

normal

force, resulting in
state

compressive strain, assuming

to an

elastic behaviour. From this strain. The tensile fraction of the

the

compressive edge is subjected thereby undergoing

a

incremental

compressive

edge

is

trial incremental tensile strain,

set to a

compressive increment

Between the two

edges,

linear strain distribution is assumed.

a

For each fibre the concrete and steel both

area

produce

force, calculated

as

their respective

within the fibre times the stress which is based

on

the strain in that

particular fibre
normal

according to the material models

The fibre forces
are

of figures 3.8 and 3.9.

cross

accumulated for the entire

section

giving a resulting

force. The initial normal force N

acting on

the

cross

section is

compared to the resulting

is checked
as

force and
follows

residual force results from subtraction. Vertical

equilibrium

%V+ %%-**%
i

(3-33)

where the first and second terms represent the concrete and steel fibre forces

respectively,

the third term the normal force, and the

right hand side

are

the force residual. Summation is

as

performed over all n fibres.

The fibre forces

obtained

F=Ae<t
F.s

(3.34)

Alal
larger than a preselected value,

(3.35)

If this residual is

the tensile strain and thus the location

of the neutral axis is

strain

adjusted,
a

and

a new

iteration is

performed, with

the

compressive multiplied
a

kept fixed,

until

sufficiently

small residual is obtained.

In addition to

summing

up the fibre vertical

forces, these forces

are

also

with the fibre distance to the centroid internal

(location of the normal force), giving

on

resulting

bending

moment which is

acting

the

cross

section due

to

the selected strain

distribution. The internal

bending moment is obtained as

/r*i + 2X*.
i
i

(3-36)

where xt is the distance from the fibre to the centroid,

as seen

in

figure 3.7.

30

CHAPTER THREE

Theoretical

curve

Simplified trilinear
curve, fitted to theo

retical

curve

8C

Sy

a) Moment curvature relation

b) Outer spring force versus axial

deformation Figure 3.10 Nonlinearflexural behaviourfor wall cross section

The curvature of the when vertical the wall section is also recorded

cross

at

the end of the iterative

procedure

equilibrium is reached, taken as the difference of the edge strains divided by

The tensile

length.

edge strain
in

is increased

by

new

increments, and in this

manner we

obtain

a moment curvature seen

relation of the wall

cross

section which is extended

into the nonlinear rnge,

This

figure 3.10a. modelling the nonlinear

The
moment

moment curvature

relation is intended for guidance when

flexural behaviour

by assigning nonlinear properties to the outer Springs.

is

transformed into
between the

couple (two spring forces) by dividing

This distance should not be chosen

the moment

by the distance

Springs.
have

significantly

smaller than the

was

length

of the prototype in order to obtain realistic

we

dynamic

behaviour. As

stated in

section 3.3,

F-M

(3.37)

It is then

possible

to

fit

bilinear
see

or

trilinear

curve

for the

spring behaviour

to the

resulting

nonlinear force relation,

Figure

3.10b. This represents the skeleton curve,

and to which the behaviour returns upon

which is followed for pure monotonic

loading,

unloading
break

and

reloading as discussed subsequently. For a trilinear simplification, the two

cracking
and

points

would represent flexural

yielding, respectively [Lind88],

[Lind89].

MACRO MODEL

31

*,A

Yield stiffness taken

as

around O.Ol K.

Cracked stiffness in tension

Elastic

stiffness in compression, Ke

Figure 3.11 Chosen skeleton

We will here

curve

forflexural Springs

simplify the approach further by assuming that the wall is to some degree
at

cracked
have
a

(flexurally)

the

beginning of the analysis.

as a

Thus due to

zero

normal force

we

cracked stiffness, which is set

fraction of the elastic

(uncracked) stiffness.
0.8

Factors of cracked to uncracked stiffness

ranging

from 0.5

to

give reasonable
to

agreement with

experimental data approximately resulting in the

procedure is more pragmatic
a moment curvature

same

ratio of cracked

uncracked flexural stiffness. This

the cracked stiffness

than
as

attempting to obtain
[Lind89], and may
the

direcy

from

relation

in

be

justified due
a

to

the fact that

essentially the moment curvature relation only gives spacing,

behaviour at
In

flexural crack
flexural cracks

reality,

only
a

occur at some

with uncracked sections in

between where the concrete has flexural

stiffening

effect. The cracked stiffness is

kept

until

yielding

oceurs.

The

yield

force level may be extracted from the transformed

moment curvature

relation. It should be noted here that the

discussed above
to

yield curvature is determined

cracked stiffness until
moment curvature

from the

procedure

by using

preselected
steel and

yielding, rather than attempting

relation. To account for
between cracks, the
some

read the

yield

curvature

from the

strain

hardening of the
is
set to some

stiffening effect of concrete

few percent

yield stiffness

fraction of the uncracked stiffness

a were

chosen

by the user.
skeleton

Values
a

ranging from a fraction of one percent to

one

tested, and generally

value of around

percent

was

found

to

be reasonable. The
as

suggested
3.11.

curve

for the outer vertical

Springs will thus be

shown in

figure

32

CHAPTER THREE

Small stiffness around

10~KCS

for numerical purposes

Sa

Elastic

stiffness in compression,

K.

es

Figure 3.12
The central vertical
moment

Behaviour

of central vertical spring

and

spring is active
cross

in

compression only,
to

consequeny
a zero

when the

acting

on

the

section is

sufficientiy large

produce

strain at the the tensile

centroid the stiffness of this

spring is

set to zero.

A small value is

assigned to

stiffness in order

to

avoid numerical difficulties. The behaviour of the central vertical

spring is

shown in

figure 3.12.

The above discussion

implies that the two outer Springs alone

The moment transferred

govern the behaviour at

large

curvatures, and that any initial normal force will be transferred

through

that outer
a

spring which is in a compressive State.

normal force

by a cross

section with

certain effective normal force N, is then

equal to the moment of a cross section with

outer

zero

plus the

normal force times the distance between

spring

and centroid,

i.e.

M"

M+

(3.38)

where MN denotes moment for

moment

a cross zero

section with normal force N, and

an

M denotes
which

for

a cross

section with

normal force. This is

approximation
to one

implies aetually
hence

that the line of action of the found

to

compressive

force is close

edge,

which is

be the

case

in walls with low to moderate normal force. With this

on

approximation
on

the influence of the normal force

the

position

of the neutral axis and

the size of the

compression area of the cracked cross section is neglected. This

to

is justified for walls with low

moderate normal force.

Generally

the above mentioned

a

expression "low

to

moderate" normal forces may be

may occur, if the

understood in such
macro

way that

no

signifieant compressive yielding

model is to

give reasonable aecuracy.

The effect of higher normal forces may for

MACRO MODEL

33

500-T(ton)

Figure 3.13 Boundary

element elongation

over

first storey versus base

shear, full-scale seven-storey wall tested at Tsukuba [ACISP84]

this purpose be counteracted

by protruding boundary elements,

in order

to

limit

com

pressive nonlinearities,

and the model will function well.

With the above the monotonic behaviour of the wall model is established. We have treated the skeleton
curve

for the outer

Springs, i.e. elastic compressive stiffness, cracked

in

tensile stiffness, and

yield stiffness,

as seen

figure
and

3.11. We will

now

determine the

unloading and reloading

Although
was we

behaviour used for cyclic

to

loading.

attempt

explain

the

unloading

reloading behaviour physically,

to

as

done for monotonic

behaviour, experimental data which pertains exacy

not

the

outer

vertical

spring discussed here is

element

available. However, certain data is available from the

tested
at

fll scale seven-storey wall

specimen
versus

Tsukuba

[ACISP84],

in the form of

boundary

elongation

base shear. If
to

we assume

that for pure static

cyclic
the

testing,
test

the base shear is

proportional
serve

the

spring

force in the outer vertical

spring,

results from Tsukuba may

in the

development of a reasonable hysteretic model,

In

describing the force-displacement relationship during cyclic behaviour. boundary

element

figure 3.13

the

elongation

over

the first storey

versus

base shear for the full-scale

seven-storey wall

example is shown.
by Kabeyasawa Springs,
al

The

original

model

et

[KSOA82]

used

relatively complicated

hysteretic

model for the outer

which is shown in

figure 3.14.

34

CHAPTER THREE

<D,.F_>

0,

Dve

=(D,-3yc)

Figure 3.14 Original hysteretic model for outer Springs by Kabeyasawa et al [KSOA82]

a) Small amplitude eycles

b) Large amplitude eycles

Yanes

Figure 3.15 Hysteretic modelfor outer Springs by

[Yane82]

yielding

in

tensien

t
(l-)h
"r0 /.in compression

TE A
c

element

E A
ss

t'NiEA/EA
c c

(s)
s

L
1

L (c)
'

.element

1 h

<c

A1
I

|eac
t"c

ui[1<y"'-i]

Figure 3.16 Modified hysteretic model for outer Springs by Vulcano

& Bertero

[VB87J

MACRO MODEL

35

Foive

Fi,

(k.

a) Small amplitude eycles

b) Large amplitude eycles

Figure 3.17 Modified hysteretic model for outer Springs by Linde [Lind89]
f

(MX

uy).

Figure 3.18 Modified hysteretic modelfor outer Springs by Fischinger et al [FVSFZD90]

Other authors modified the

hysteretic
3.15

models for the outer

Springs

and

proposals for
on

modifications
these models

are

shown in

figures

to

3.18. For notation and detailed discussions

see

the

appropriate references.
were

All the models shown hitherto

where the

mainly used in static cyclic numerical applications,

obtained from
some test

prescribed deformations
wall tested

were

program, such
or

as

the

full-scale seven-storey wall

models of the
same

structure
at

tested in Tsukuba

[ACISP84]

other reduced scale

same

different laboratories within the framework of the

was

research program. The fll scale wall structure

tested

by the SPD method (Single

degree of freedom Pseudo Dynamic test)

which
were

which resulted in floor

displacement

histories

subsequently used in numerical analyses.

36

CHAPTER THREE

The

macro

model

developed

in this report is to be used with

general
the

finite element

code, the dynamic analysis being performed by direct integration methods using general

ground

motion histories
macro

as

input Therefore,
are

the

requirements

on

hysteretic

model

used with this

model

simplieity and reliability

for global results.

Simplieity is needed mainly due to the already heavy computational demands by the
method of

analysis.

It is not desirable to combine

very

complicated hysteretic
method.

model every

having

many different stiffness branches with this

cause

analysis
models

Generally,

change of stiffness will

consumes

residual forces
time. The

requiring equilibrium

iteration which
some

costiy computational

hysteretic
by

suggested by
same

authors,

and shown above, often

display several

stiffness branches in the involved

direction, and the

In

location of stiffness break

was
a

points

are

determined

expressions.

[Lind89] it

shown that relatively

simple hysteretic models may in general give at least as reliable

there should

global result with regard to test data. Essentially justified physically

or

only be a stiffness change

when it may be

is

clearly

necessary from

experimental

data.

Generally, the fewer the stiffness changes of the model the better. Despite the simplieity requirement discussed above
be reliable
more

good hysteretic

model must still

compared to experimental data.

a

It may, therefore, not be possible to

some

simplify it

than to

certain

degree. Otherwise,

of the characteristic behaviour of

structural walls may be lost or

impaired.
the outer

We propose here

hysteretic model for

figure 3.17,
region

Springs,

seen

in

figure 3.19,

which It

major characteristics are simplieity and reasonable agreement with experimental results.
is based break
on

the model in

and is made

more

efficient

as

follows. The stiffness

point

in the tensile

of the skeleton

curve

is moved to the

origin,

i.e. it is

assumed that the wall is

already slighy

cracked. This does

assumption
a

is

quite

realistic for

existing buildings,
construction due induced
or

since

slight cracking

usually exist
as

number of years after

to

wind and other actions such

In this way

environment and

possibly traffic-

seismic

ground motion.

some

input data for the

efficient.

tensile crack

point

may be omitted and the numerical behaviour may be

more

Secondly, the proposed

figure 3.19b.
the elastic of All

model

displays

modified

reloading behaviour as

seen

from

reloading
a

oceurs

direcy

towards the maximum tensile

displacement
including
as

reached. In this way

purely trilinear loading

where cracks
four

and reloading model is obtained

compressive branch
3.17

are

closed. Previous models such

which is rather
a

the

one

figure

displayed

reloading branches

lot The

cyclic

behaviour

using this simple and effective hysteretic model is

shown in

chapter five to be

MACRO MODEL

37

a) Small amplitude eycles

b) Large amplitude eycles

Figure 3.19 Proposed simplified hysteretic model for outer Springs

quite satisfactory
mental data. The

with

regard

to

the overall behaviour of the wall

compared

to

experi

proposed
versus

model in

figure

3.19

displays

the force in

one

of the outer vertical

or

Springs

the axial deformation in that

spring.
to

The
an

loading

"skeleton"

curve

consists of three basic stiffhesses. It is modelled with

which is used when the wall is first
and later

elastic

compressive stiffness Ke
terms

subjected

gravity

load in

of normal force,
into
a

during cyclic
certain

behaviour when

one

edge (one

outer
cross

spring) gets

compressive State.
to

In tension the

simplification
prior to
the skeleton

is used that the

as

section is considered

be cracked to

extent

analysis
curve.

discussed above, thus

giving

cracked stiffness Kcr

as a

already for the

one

Finally the yield stiffness Ky is taken

fraction of around

percent of the elastic stiffness.

shown in the
in
a

Behaviour
In tension

prior to yielding is

figure on the left (small amplitude eycles).

a

unloading always
branch

oceurs

direction towards after

point

on

the elastic

compressive
The

representing crack closure

right displays

compressive yielding of the

flexural

reinforcement steel.

figure on
the

the

the behaviour upon

yielding (large amplitude eycles).

Ku, parallel
to

The

unloading
at

from

yielding

oceurs

with the stiffness

the

unloading
From

stiffness there the

and the

yielding point
section

as seen

in the

figure ("elastic" unloading stiffness). compressive

elastic branch, chosen

cross

yields in compression (reinforcing bars bridging open cracks yield)

point
onthe

unloading

moves toa

bythe

38

CHAPTER THREE

Figure

3.20

Hysteretic rule numbers for suggested model

user,

which represents the

on

point

where cracks

are

closing.

This

point

was a

found

to

be

dependent

the normal force, and is set to the force tends to

yield force multiplied by

factor aci. A order

larger normal
to

require

larger negative force level for crack closure in

to

obtain realistic

global hysteretic behaviour, compared

experimental

data. A force

level in the rnge of

-Fy was found

as

realistic for small normal forces.

For

high

normal forces,

found in the

case

of many storeys and

high gravity loads,

of

the crack closure force level tends to increase to e.g.

-AFy

for the

case

eight storeys

(normal force
a

on

the wall at first storey

equal

to

MN). This is explained by the fact that

large part of

the internal moment is made up of the axial force

(passing through
more

the

compressive
internal
at

zone

located

near

the

edge).

This part may account for

than half of the

moment

in

multi-storey

walls. The tensile forces

a

(flexural reinforcing bars mainly

the tensile

edge) only

contribute

smaller part ofthe total internal moment

Reloading
maximum and

oceurs

towards
was

point

on

the

yield

branch of the skeleton

in the

curve

where the

displacement

reached earlier,

as seen

figure.

For

partial reloading change

at zero

unloading

the stiffness is

proposed
a

shown in the

figure,

with stiffness

force level. In order to obtain

clearer overview of the

hysteretic rules,

it is useful to
are

assign

number to each rule, and

are

or

branch, in figure 3.19. These numbers

as

shown in

figure 3.20

briefly

described

follows:

MACRO MODEL

39

Rule 1. Elastic At the

compression.
the element is

beginning of the analysis

expected to be at rule 1, due to the effect

tensile stress, the rule

of gravity load. Due to the

assumption of some cracking prior to the analysis no tensile reaching changes

to

elastic stiffness is considered, and when

rule 2.

Rule 2. Cracked tensile stiffness.

Is entered from rule 1. Models the cracked tensile stiffness without

prior yielding.

Unloading leads to rule 4. Further tensile force leads to rule 3 representing yielding.
Rule 3.

Yielding.

Is entered from rule 2 if no

rule 7.

prior yielding.

If prior

yielding occurred, it is entered from

Unloading

leads to rule 5.

Rule 4.

Unloading from cracking.

Is entered from rule 2. Below force level

entrance

-aciFy it

leads

to

rule 1. At forces above

level, it leads back to rule 2.

Rule 5.

Unloading from vielding.

zero

Is entered from rule 3. Below

entrance

force level it leads

to

rule 6. At force levels above

level, it leads back to rule 3.

Rule 6.

Unloading at negative force level. yielding,

force level. At force levels below
to

Is entered from rule 6. If reloading has occurred which did not lead to further it may be entered from rale 9 at
to
zero

-aciFy it leads

rule 8. If reloading oceurs, it leads

rule 7.

Rule 7.

Reloading.
or

Is entered from rule 8,

from rule 7. Takes direction towards maximum rule 3. If unloading oceurs, it leads to rule 9.

displacement

reached where it leads

to

Rule 8.

Compressive elastic

stiffness.

Is entered from rule 6. Force levels above

-aciFy lead to rule 7.

Rule 9.

Unloading from reloading.

to

Is entered from rule 7. If force level is below zero, it leads

above entrance level, it leads back to rule 7.

rule 6. If force level is

40

CHAPTER THREE

4.0
.

"

""

-*
.

""""""

r-

3.0

/'r

2.0

fi

// /''

E
o

fi
li
1.0

aer aer

0.8, ay 0.5, ay

0.02 0.01

fi
Tf

0.0025

0.005

0.0075

0.01

Fourth floor horizontol

displacement [m]

Figure 321 Influence of outer vertical spring stiffness on global flexural behaviour
The model

suggested here for the hysteretic

an

behaviour ofthe outer vertical

macro

spring

was

found to result in

overall flexural behaviour for the

model which

appeared

reasonable

compared to experimental data.

of this

It should be noted that the appearance of the

more

hysteretic loops
such

spring

is in itself of

limited value,

although

some

comparison with experiments could

as

be made of its deformation

versus some outer

force

the base shear. The vertical

spring

is otherwise

essentially artificial
the

in order to

reproduce

the entire flexural behaviour.

the

Concerning
Partly

resulting
to

overall flexural in

hysteretic behaviour,
was

original

vertical

spring hysteresis model,

this appears

seen

figure 3.14,

found

to

give

thin

loops

sometimes.
on

be due to the

unloading stiffness

which is
a

dependent degree

the

maximum

displacement reached. This

of the model

has been observed to

small

from

some

experiments, although usually not as pronouneed as suggested in the original model.

unloading stiffness
constant,

The

proposed

in this report

was

deliberately

chosen to be

partly

due to observed behaviour and

some

partly for

the sake of in the

simplieity.
model

The
was

variable location of

of the stiffness break

points

original

deliberately abandoned,
rules
were

which did not

to

impair the overall behaviour. The other hysteretic

be
as

also

deliberately chosen
a

simple

as

possible,

but

as

comprehensive general.

as

necessary to obtain

reasonable Simulation of flexural wall behaviour in

Before

turning

to

the behaviour in shear

some

remarks

are

made

regarding the relation

and the

between the "local" behaviour of the flexural

Springs described here,

"global"

MACRO MODEL

41

-imiiiitttimmtttmnRm

Ke(Kcr> fy)

Ke

*?&*.%./>>

Hat.

tt

Y*\YY7,^afr
,

Ka

rg

***>tm.*t-i*ittiti*rr*m*-

\-

IrSection

Figure

3.22 Moment

of inertia made

up

oftwo spring areas (spring stiffhesses)

flexural behaviour. We limit the discussion here to the skeleton curve, i.e. to the rules 1
to

3 of figure 3.20,

comprising

uncracked stiffness, cracked stiffness, and

yield stiffness.

In order to show this

relation,

simple example is
section with
a

shown in

figure 3.21, displaying

of 3.0 MNm is
shown

flexural behaviour in the form of a moment versus curvature result

plot obtained as a numerical

moment

by

the
to

macro

model. A wall As
a

cross

yield

subjected

pure

bending.

reference,

completely elastic behaviour is Springs

by the
the

solid line. For this reference the tense stiffness of the outer
uncracked

was set

equal to

compressive stiffness,
obtained, each with

and
a

no

yielding

was

allowed for. Two

comparison
of the outer

solutions
vertical

were

value for cracked and

yielded stiffness
as

Springs.

The first Solution set the

spring

stiffnesses

follows: cracked: 80 % of

uncracked, yielded: 2 % of uncracked. The second Solution had the stiffhesses: cracked:
50 % of uncracked,
The

yielded:

1 % of uncracked.

global flexural behaviour resulting

from these solutions may be measured in

as

figure

3.21, in the form of flexural stiffness defined

moment

divided

by

curvature.

Normalising
the first

the elastic flexural stiffness of the reference Solution to

unity,

we

obtain for

comparison

Solution

global

flexural cracked stiffness of around 89 % and

we

global

flexural

yield

stiffness of around 4 %. For the second Solution

obtain

corresponding

values of around 65 % and 2 %.

These values may be confirmed wall model is shown each other at

a
a

by solving

this relation

analytically.

In

figure

3.22

consisting of two Springs with different stiffnesses, separated

only
deal here with the
case

from

distance /. We

of two active

Springs, assuming alternatively, spring does

that
not

deformed State such that the central vertical have

a

spring

has

"yielded",

or,

we

completely elastic compressive

State, in which the central

contribute to the flexural stiffness. For the discussion of the wall section moment of

42

CHAPTER THREE

inertia

we

work here with the

areas

of the

outer

Springs,
area

rather than with the

spring

stiffnesses. Since the

spring

is

an

artificial creation its

is

proportional to
an

its stiffness

(height h,

and concrete modulus E remain

to

unchanged).

We define

idealised

spring area
and
area

Ke, taken equal

unit E-modulus, is

the elastic

spring

stiffness

A I h assuming unit element height,

the stiffness is

see

equation 3.7. Each time

two

changed,

the

spring

A.

changed,

since the other

factors in the stiffness

expression

are constant.

The

moment

of inertia about the strong axis of the

of Steiner's rule, be written

completely elastic model, Ie,

may then,

with

use

Modifying one of the two idealised spring areas to the cracked value Km.,
moved centroid, located from the

we

obtain

right edge a distance lc, where

lc

Kerl

Kcr + K. corresponding global cracked moment of inertia

(3.40)

and the

L^K^l-lmf + Kmilmf
It is
now

(3.41)

possible

to define factors of reduced

global

moment

of inertia divided

by

elastic

global moment of inertia, i.e., for cracked idealised

area, and

thereby for cracked

stiffness

ot

(3.42)

and

correspondingly for yield stiffness

L

a'=-f
i.
t

(3.43)

the latter

basing

on

expressions (3.40)

and

(3.41)

with the cracked

spring

stiffness

Ka

substituted

by the yield spring stiffness Ky.

MACRO MODEL

A3

1
1
a

1
0 0.2
Local

0.4

0.6

0.8

1.0

(spring)

stiffness

ieduction,a<r ,a

Figure 323

Relation between local

(spring) and global flexural stiffness

reduetion

Inserting the spring stiffness values

a

ofthe numerical

example

shown in

figure 3.21,
to

as

fraction of the

to

unity normalised elastic spring stiffness, into expressions 3.40

3.43

we

obtain for the first comparison Solution

=

cracked stiffness, (Xcr

0.8:

le =0.444/ le =0.0196/

and and

c= 0.889

yield stiffness, Oy
and for the second

0.02:

a'y= 0.039

comparison Solution
=

cracked stiffness, Oa-

0.5:

lc= 0.333/
lc= 0.0099/

and

=0.667

yield stiffness, Oy
all of which show

0.01:

and =0.020

good agreement
for local
be

with the numerical result As

rule of thumb it is

possible to
the the local
to

State that

(spring) yield stiffness in the typical rnge of a few percent,

roughly twice that of the yield spring stiffness.
stiffness
to

global flexural stiffness will

As for

(spring) cracked stiffness, one may estimate the global

be

roughly

10
a^

25 percent

higher than

the

spring stiffness,

with the

higher value applying to lower

the

values. A continuous relation between the local

(spring) and

global (flexural)

stiffness is

shown in stiffness

figure 3.23. On the

horizontal axis is shown the reduetion of the

spring

(local), and the corresponding global flexural stiffness reduetion is plotted

corner

against the vertical axis. The region of the lower left

represents the

of the

figure typically

yield behaviour, whereas ie

rest

of the

figure essentially represents

cracked behaviour. The upper

uncracked behaviour.

right corner with

ccordinates (1.0,1.0) represents elastic

44

CHAPTER THREE

3.6 Shear Behaviour

The behaviour in shear is

completely

simulated

by

die horizontal of this

spring.

In accordance

with earlier discussion,

we

derive here the

properties
theory,

spring by comparing
in

model

behaviour to that of

real wall

obeying

elastic

as seen

figure

3.24.

'/////////////////;
a) Elastic theory

b) Model

Figure 3.24 Shear behaviour

For elastic behaviour obtain

we

by elastic theory the horizontal deflection of the real

wall

as

follows

8*.=

KVh

(3.44)

where G is the modulus of rigidity, and

is the form factor in shear, which is geometry-

dependent

and often ranges from 1.2 to 1.3. Tomii and Osaki

[T055] give the following

empirical value

of

x*

K_

0.75(1 + u)[l-u2(1-v)

[l-3(l-v)]
geometrical parameters
for
a

(3.45)

where the

and

v are

shown in

figure 3.25. The elastic

shear

stiffness

Kve

section of height h is then

given by

-GK

K...=

hK

(3.46)

MACRO MODEL

45

l'=uL

l=(l+u)L/2

Figure 3.25 Geometrical properties forformfactor in

This stiffness is used until shear

shear

cracking

oceurs,

whereupon

which

cracked

stiffness is stiffness

employed.

An

empirically-based
Otani et al

ratio a. of cracked to uncracked shear

as

was

suggested by

[ACISP84]

follows

,
'

0^ + 0.14
fc

(3.47)

where pw is the horizontal reinforcement ratio.

ment
cross

Equation

3.47

gives

for

typical

reinforce

ratios

factor as in the rnge 0.15

to

0.20. The parameter as

was

derived for wall

sections of different geometry

ranging from rectangular walls (f

b)

to walls with

pronouneed protruding boundary elements.

The

cracking

force in shear,

Vc, may be estimated quite simply by

(3.48)

VC=Kfrwhere Am is the the shear

cross

sectional

area

of the wall, inclined at about 45

degrees,
a

and / is

strength

of the concrete. It should be noted that

to

(3.48) gives

very

simplified
more

result, without any regard

normal force, geometry, reinforcement etc. A

was

comprehensive empirical relation ^c

=

suggested by Hiraishi [ACISP84]

as

W.
pJs

(3.49)

where

Kt is

form factor

given by

46

CHAPTER THREE

1.05t/b

(350

where shear

is the thickness of the web,

lp

is the

length

of the web, and f. is

an

equivalent

stress

given

as

/*asCfV1+c"
where cr0 is axial
stress

(3"51)

and Q is

given by

C,

=2.2>/J-0.00546/;1-456pw007S
3.49
to

(3.52)

Equations
For
a

3.52

are

taken

direcy from the

source

and

use

units

kg

and

cm.

simplest possible estimation of the ultimate shear capacity we may add the concrete contribution of (3.48) to the horizontal steel contribution of the bars with area A^ crossing the inclined surface,
as

suggested in [Lind88]
(3.53)

Vm

V + V,=AmJ, + AJ,
realistic estimation is

more

provided by

the

empirical expression of Hirosawa

[Hiro75]
0.0679a023C+17.6)

"
,

v*

_.

r-

+
m

VM/VX

+ 0.12

0.845^/^0^MJrryH^k + 0.lo"0o

A.

bJ

(3.54)

where the result is

given in MPa,

and the constituents

are

explained as follows

p,

effective tensile reinforcement ratio in percent

"=W-tii)
A.
=

(3-55)

area

of vertical reinforcement in tension side shear span to

boundary element,

MI(VL)
f
=

depth ratio,

yield strength of horizontal wall reinforcement

MACRO MODEL

AI

p^

effective horizontal wall reinforcement ratio

P~+=TL

bts
s
=

(3.56)

vertical
=

spacing of horizontal wall reinforcement

over

o.

average axial stress

entire wall

cross

section

area

b.

average width of wall section

j, L and h are geometrical parameters of which

HH)
and forL and h,
see

(3.57)

figure

3.24

Since the

yielding and ultimate strength levels

used
a as

in shear

are

relatively close, the

small fraction of

expressions (3.53) or (3.54) may be

as was

the

yield level in a trilinear hysteretic model,

as a

done in

[Lind89], followed by

yield

shear stiffness taken

the uncracked shear stiffness. However,

relatively
in
a

little is known about this behaviour

generally, and since

is

shear failures often

occur

brite manner, this

region of behaviour
here.

clearly

to

be avoided, and wl not be treated in the model

are

developed
in the

Design

recommendations for avoidance of shear failures

provided

capacity design

method,

see.

e.g.

[PBM90]

or

rPP92].

Concerning
behaviour for

the

cyclic behaviour less weight

still

is put here

on

modelling

the

unloading
An

large

shear deformations, due to the abovementioned

reason.

experimentally obtained curve is

on a

presented in figure 3.26, obtained from cyclic tests

tested
at

three-storey wall specimen [VBP79]

the

University
versus

of California,

Berkeley. during

It shows horizontal shear deformation at the third floor shear deformations

are

base shear. Since these

seen

relatively large, the typical pinching behaviour may be

unloading, figure

discussed further in [VBP79]. For smaller

eycles

it may be

seen

from the

that the

pinching is
not

mueh less

pronouneed.

It should be further noted that the test

specimen was

capacity designed and the test programme did not correspond to actions

expected for a capacity designed structural wall.

The

original

model

by Kabeyasawa

et

al

[KSOA82] employed
was

trilinear

origin-

oriented

hysteresis model for the shear behaviour, which

modelled

by the shear

spring

alone. The

loading curve is made up of a trilinear skeleton curve, consisting of an

48

CHAPTER THREE

Figure 326
wall

Shear deformation
tested
at

at

thirdfloor versus base shear,

three-storey

specimen,

the

University of California, Berkeley [VBP79]

elastic part, towards the

curve

cracked part, and

yield part.

The

unloading
3.27.

curves are

always directed

origin.

The

reloading follows the

latest unloading

path until the skeleton

is reached. This trilinear model is shown in

seen

figure

As may be

from

figure 3.27, the origin-oriented model does

not

give

particularly

realistic Simulation of

typical

shear behaviour for

to

large deformations.

The

loops

are too

thin in the

vicinity

of the

origin compared
be

experimental results, and,

Rather it should of the

furthermore, the unloading path should

not

direcy origin-oriented.
some

produce plastic deformations

and

possibly,

in addition, exhibit

pinching

(slip) behaviour typical for shear.

On the other hand, with behaviour

design according

to

the

capacity design detailing)

method

(nonlinear

predominany

of flexural nature, and careful

we

do not enter into and is of less

regions

of excessive shear deformation. Mueh of the

pinching disappears

importance to model.
are not

For the

analysis of general buildings, i.e. including buildings which regard the possibility of strong
and exhibits

capacity designed,

it may be necessary however to

nonlinear shear behaviour of walls.

A modified shear model
was

proposed by Fajfar et al,

the

elasto-plastic slip proposal,

Colotti

behaviour. This model is

seen

in

figure 3.28, with elastic, fully-plastic behaviour on the

on

left, and elastic, linear strain hardening

right

In

different

[Colo93] receny attempted to calibrate the shear spring by

means

of a shear panel model.

MACRO MODEL

49

Figure 327 Origin-oriented trilinear modelfor shear by Kabeyasawa

et

al

[KSOA82J

Q
F
i

/
/
'

/
/

/
/

"!
J

Qy-

/^77
rr

^
A

^^

aj Afode/ suggested by Fajfar

and Fischinger [FF90J

b) Model suggested by Fajfar et al [FVF92]

Figure 328 Hysteretic slip models suggestedfor shear behaviour

An

improved hysteretic

model

was

proposed

in

[Lind89] which is shown in figure

3.29. The model attempts to

closely simulate the experimental behaviour of concrete

a

under shear. This, however, is stiffnesses.

relatively complex model employing several unloading

50

CHAPTER THREE

Force

f,

f~

Yr~"

F-

-rfjfjfmjr

j /i.

Dcfocn

flj Small amplitude eycles

b) Large amplitude eycles

Figure 329 Comprehensive trilinear hysteretic shear behaviour model [Und89]

The smdies in this report

mainly

deal with structural walls which behave

predomiserve

nantiy in
as

flexural mode. The models

developed in

this report, furthermore,

mainly

numerical tools
are

together

with the

capacity design method,

in which extensive shear

deformations
Based
on

avoided.

these

assumptions

it is considered suffieient to this report. Thus

we

use a
a

bilinear model for the

horizontal with
an

spring for the smdies in

a

obtain

bilinear shear behaviour,

elastic stiffness, and

cracked stiffness in each direction. The bilinear

to

hysteretic
relatively
designed

model is shown in
few stiffness

figure 3.30. This model is numerically efficient due

its

changes,

and the shear behaviour which is of interest for capacity

walls is modelled

quite realistically.

MACRO MODEL

51

Figure 330
For the

Bilinear shear behaviour model used in this study

cracking force of the bilinear model

we use

equation (3.49)
and used

or

alternatively
limit on the

(3.48), and for the cracked stiffness equation (3.47),

models. The ultimate force of (3.48) may be

as was

intended for the trilinear

as a

readily calculated
not

cracked

path of die binear model, which should

be exceeded

during

numerical

analysis.

3.7 Element Formulation

The

macro

model

developed

in the

previous
and

sections of this

chapter

is to be used in

general finite element code,

behaviour,

see

[Abaq89]
general

[Abaq91]. All the physical properties deve

axial behaviour and shear

loped in the previous sections, comprising flexural behaviour,

are

introduced into
over a

wall element intended to simulate

completely

typical

wall behaviour
to a macro

chosen

height

of

structural wall. This element will be

referred

element

The

generally formulated

element

equation

of motion

(3.58) contains three

to:

terms on

the left hand side,

inertia.
In order to
case

representing

internal element forces due

stiffness, damping, and

distinguish between internal

to

and external element

properties we use lower

case

letters for terms related within the

macro

the

macro

element and upper

letters for

spring
and

properties

element Vectors and matrices will be denoted

by bold

52

CHAPTER THREE

scalar

properties by

normal letters. The summed element forces may then be written

as

follows

f,+fd+f.=fg
where the terms
are

(3.58)

given by
(3.58a)
(3.58b)

f,=ku fd=c

/.

(3.58c) (3.58d)
tn

ft=ag(t)m
where k, c, and

represent the element stiffness,

damping, and mass matrices and, ,

and acceleration vectors. The

and represent the element

displacement velocity
the

right
of

hand side

dynamic force acting on the element's degrees freedom, where ag(t) represents the general ground acceleration.
In this section
we concentrate on

expression represents

the formulation of the element stiffness matrix

Kt.

In

chapter five the modelling of the other two internal forces is discussed.
The nonlinear stiffness the

properties described in the preceding sections

model shown in

are

adopted

to

Springs
a

of the

macro

figure 3.3. The introduction of this

three different versions. the separate

macro

model into

usable finite element was

performed in

The first and

were

simplest version involved

assembly

of the

Springs

which

modelled

as

2-degree

of freedom

(d.o.f.) elements,

and further
was

by

the

assembly of

the stiff beams modelled

as

6-d.o.f. elements. This version

mainly

used for initial

kinematic verification tests, The second version into

one

involving one or a few storeys.

assemble the

adopted is to numerically

Springs

and the stiff beams

"macro element". In this

some

Option

number of undesirable

degrees of freedom
static condensation.
and allows for

appeared,
Apart

of which may be
macro

relatively easy Condensed out by

at

from

that, the

element in this version is

relatively versatile,

connections of girders and joining walls

mid-height of the web.

a

The third and final formulation involves

derivation of

a macro

element based

on

linear kinematic relations. In this way the flexural stiffness of the

omitted and
a

rigid beams

may be
an

stiffness matrix may be obtained which has

closed form. Thus,

see

element is obtained which

essentially

acts as a

membrane element,

figure 3.31,

and

MACRO MODEL

53

Ut

Uk9
vv"vWWWl

M6

I ~~Miimwvi'.w

lort^fs

u.
*35&&asa;m3i^^

^2

t*

I Jh
b) Macro element

a) Spring model behind macro element

Figure
which be

3.31 Macro wall element with ten

degrees offreedom

can

readily used for the numerical modelling of multi-storey structural walls. figure 3.31
ten

On the left in

is shown the

macro

element which
are

was

programmed

and

tested in this report It has

d.o.f.'s of which two

rotational for connection to frame

as

girders.

The stiffness
to

properties of this
of freedom,

element
are

expressed

the stiffness terms

corresponding according to
the left in

the

degrees

based upon fictitious

Springs arranged
shown
on

the

previously described

nonlinear

spring model (macro model),

figure 3.31.
macro

The
the

resulting

element may be used

essentially

as a

membrane element,

seen on

right in figure 3.31,

ten

but offers the convenient connection

possibility for girders at the

with
a

top. The

d.o.f.

macro

element

as

implemented may be used

,

few

input data only

None of the

pertaining
internal

to

the wall

cross

section

reinforcement and yield

as

moment.

spring properties

need to be

given

input
spring
model

By expressing rigid

the internal d.o.f.'s in the

(e.g.

at

the intersection of the

ten

beams and at the central in

Springs)

in terms of the outer

d.o.f.'s which

are

displayed

figure 3.31,

we

omit excessive terms which would

normally

not

be

connected to any other element.

In order to derive the stiffness terms,

we

apply the direct stiffness method, setting the

and at the
same

displacement
displacements
are

of each d.o.f.
to zero.

equal

to

unity,

time

keeping
a State

all the other

The forces at all d.o.f's needed to maintain

of

equilibrium
to

the stiffness terms in the column of the stiffness

matrix, which correspond

the

54

CHAPTER THREE

d.o.f. for which the

displacement

was set to

unity. unity.

We thus have

ky,

the column of

stiffness terms due to the We

displacement uj,

set to

begin the

derivation ofthe stiffness matrix,

are

considering only the eight translational

In order not to

d.oi.'s. The two rotational d.o.f.'s

treated

separately later.
are

change

the

final d.o.f.

numbering both rotational

d.o.f.'s

shown from the

begirining.

We denote the stiffness of the horizontal

outer

spring by Ks,

the stiffness of the left vertical

spring by K\,

the

right

outer

spring by K%

the central vertical

spring by 3,

and

the axial stiffness of the horizontal stiff beams

by Kapreviously described
"rigid",
are

The beams

connecting the Springs,

which
as

were

as

are

for

the numerical formulation considered

follows. AU the internal beams

are

flexurally

rigid.

The vertical beams, in the centre of the element,

was a

axially rigid. Since a complete

such
an

frame-wall interaction

goal of the element capability, although degrees

of freedom at each
a

application
macro

is not treated in this report, horizontal element

are

corner

of the

needed. The horizontal beams, which have

horizontal

degree of freedom
chosen

at

each end,
to

are

provided with a large axial

stiffness Ka, which

numerically is
wall.

so as

physically represent the elastic

As

behaviour of the

tributary part of the

number

an

example we set the displacement of the d.o.f.

one

equal

to

unity,

i.e. u\ of

1,

as

shown in

figure 3.32.
macro

The stiffness terms arise from the internal

spring forces

the disassembled

element shown in figures 3.33a-c.

Figure

3.32

Deformed State atu\--\

MACRO MODEL

55

*A

Figure 333a Internalforces acting on

due
to

lower assembly

horizontal spring force, u\

-wmm

KAD

KAD
Axial forces

Figure 333b

acting

on

lower

assembly,

u\

-&?) \
rl

-bi*?:
K.-

K.

Figure 333c Internal forces acting on upper assembly,

Thus for ui

ui

1,

we

obtain the stiffness

terms

*h! *a

rr

"-r

^f(
*n=0
.
%
=

_Ksc
2 l
rr

*S1
k41

K.

~Ka+ A

2 /

k5l

*io,i=0

(3.59)

56

CHAPTER THREE

It is also shown how the stiffness terms for the

case

when 2

are

obtained, since
and the

this involves the vertical

Springs.

The deformed State is shown in

case

figure 3.34,

spring

forces and reaction forces for the

when 2

are

shown in

figures 3.35a-h.

2=1

Figure

3.34

Deformed State

at 2

I(1)

Figure

3.35a Internal

spring forces acting on

lower

assembly, 2=1

Figure 3.35b Spring force from left outer spring

and its reaction
on

lower

assembly, 2=1

Figure
spring

3.35c

Spring force from central vertical

on

and its reactions

lower

assembly,, 2=1

MACRO MODEL

57

Figure 335d Spring force form horizontal spring and

its reactions
on

lower assembly,

2=1

I* (1)
K S-

K
Figure 335e
upper

Internal springforces

acting on

assembly, 2=1

JC (1)
f (1)

FzgKre 5 35f Spring forcefrom left outer spring

and its reaction
on

upper assembly,

2=1

58

CHAPTER THREE

^-4
m

*4 f

K
Figure 335g Spring force from central vertical spring
and its reactions,
on

upper

assembly, 2=1

iii
K.
c

h-c\
l

)
1

( cY h-c

H7A/JJ
K.
c

JWMH1H1MUUIBB

2 /

2 /

Figure 335h Spring forcefrom horizontal spring and

its reactions
on

upper assembly,

2=1

We thus obtain the stiffness terms in the column

where it should be noted that several terms
shown in
are

corresponding

to 2

as

follows,
cases

created

by superposition

of die

figures

3.35a-h.

^~

fr

Kj.9.
2 /

I-

K'

**"

2 /

*52=-^y

(3.60)

MACRO MODEL

59

All other stiffness terms for horizontal and vertical

degrees of freedom

may then be derived

obtained in

similar way

as

for the

two cases

shown. Each column and the relation

was

by

setting the corresponding displacement equal to zero,

kij

(3.61)

which

applies due to symmetry of the stiffness matrix, was used rather as control. degrees of freedom,
seen

The two rotational

in

figure 3.31, still need to

be treated. We

employ
7 and

an

internal flexural beam connected

a

only

between the d.o.f.'s 7 and 10. This

internal beam has

total of four d.o.f.'s, and is assembled into the

are

degrees

of freedom 6,

9,10. The properties of the internal beam

in the

only

relevant for the connection to

joimng girders
calibrated
to

analysis

of frame-wall structures. These

properties

may be

simulate the elastic connection into the wall

were

by girders.

In this report,

only

isolated walls modelled

studied and

no

such calibration

was

performed.
of

Isolated walls may be beam. It should be elastic

numerically and analysed satisfactorily without ie internal

noted that there would also be other

possibities

modelling the

girder

connection, e.g. by rotational Springs

For the sake of completeness the

or

other arrangements. of the internal beam

are

general formulation
macro

properties

and the inclusion if its stiffness terms in the beam is shown in

element

treated briefly. The internal

to

figure

3.36 with its do.f.'s,

as

they

refer

the

macro

element of
in

figure 3.31.
book.

The stiffness matrix,

No derivation is

kn,, for

the internal beam of

figure 3.36 is given

equation 3.62.
text

given here,

but it may be found in any structural matrix

7^
Figure 336

"io/'Tn

Internal beam of macro elementfor elastic girder connection

12EI

6EI

Y2EI

6EI

/3

l2
AEI

/3
6EI

l2
2EI

kib

l
=

l2
\2EI

6EI

(3.62)

/3

l2
AEI

Sym

60

CHAPTER THREE

The blown up stiffness matrix of the internal beam, i.e. (3.62) assembled into the
macro

element d.o.f. numbers of figure 3.31, will then look

as

follows

12EI

67
0 AEI

127

6EI

/3

~~7~
6EI
~-T

im

~T

~T~
0 6EI

0
127

Sym

~7~

~~F
ah

~T

Blown up

stiffness matrixfor internal beam of macro element

The blown up stiffness matrix for the

macro

element with

only translational d.o.f.'s

is

shown

on

the

next

page, followed

by ie complete stiffness matrix including the

contributions from the internal beam with rotational d.o.f.'s.

Ks
K.

h.1
K,(h-c) K,(h-c)
'

K.
2
/
2

2 l
2

/ l

-KA+0

~^K3
'

5
s
o

*i

2 /

*+*.[j
4f
^(c)(h-c)
-~K-{i)
2 l
~

7
Ks(h-c)
2 l

^-<tJ
-4--(^
0
~
'

Ks

5l

5l
4

2
l

**
+

K'[l
2
l

0
2
l

-^--r

A-j

(cV/i-c)
K'\T\
KAh-c)
/
l

t-k{t)
K,(h-c)
2 /

'K2~T

K+-K*
+

Ks

i-

Ki
0
-

h-c\2
2

K, (h

c)
l

K,

rrfh-c)2

Ks
Ka+ A

Ks(h-c)
4

Sym

K*

A-3

(h-c\2

T+K{)

Blown up stiffness matrix for lower and upper assemblies of macro element

comprising translational degrees

of freedom

OS

K,
K'C

K,

A
.

m,K'
K.
4 l
4

K,(h-c)
2 2
l

TT
K,
c

0
0

K,

K. (h-c)

as

Ksc K.c
~

K3

(c\2
"TT
2

K>+T+K{jj
-.-''*<7)V
0

T
KAh-c)
2
l

Ka

Kt
K.c

K.
4 / 4 2

JC, (A-c)
0

Ks

"TT
*3
c

K2
-2-rK,

(tJ
K, K3
4

"TT
TT
K, K,
a

(cMh-c) K'{l) l
0

A-,
c

-<>-'; *<W'
KAh-c)
2 /

KAh-c)
2 l

K,

(h-cY
\2EI 6 EI

KAh-c)
2 /

12EI

6EI

(h-c}2 { l )
6EI

1
2E1

4 EI

~T~

~T
K,

KAh-c)
2
/

Sym

Ki

+ K.

f/i-cV
Kj-r-f

12EI

6EI

J+

f
AEI

1
~T
.

1
including rotational degrees of freedom

Complete

stiffness matrix of macro element

MACRO MODEL

63

3.8

Ductility

demand

The

monitoring of the ductity

of certain

preselected

critical

zones

of a structure is The

one

of is

the most

important
in

observations to be made

during

nonlinear
on

analysis.

designer

then able to compare this behaviour with the

assumptions

which the

design was
to take

based.

Especially

plastic hinge
detailing

zones, where flexural

yielding is intended

place, the
this

demand for curvature

and the structural

ductility
must

needs to be known, since the

design of the wall section

on

allow for this demand. More elaborate discussion

topic is provided by Paulay et al in [PBM90].

The above described Macro model will

readily

lend itself to the

computation of

ductility values during given by

nonlinear time

history analyses.
definitions
as

In order to discuss the

ductility
is

measurement we will need to

make

some

follows. Rotation 0, in

general,

Q=Svl~ Svi
l where

(3.63)

8vl

and

8^

are

the vertical

displacements of the
in
one

outer

Springs. We
as

define the wall

see

rotation at which

yielding begins

of the outer

Springs

the

yield rotation,

figure

3.37. The rotation for which the left outer

spring just yields,

i.e. at the onset of

yielding, is
8^-8

e>=~i~

(3.64)

l^j^j^^^^^^lfe^h?*lft'ifc^>?1li!^.>>jl-l

d) Rotational ductility

b) Curvature ductility

Figure 3.37 Ductility measures

64

CHAPTER THREE

For rotations rotational

larger

than the

yield rotation

the ratio of 6 to

dy

is defined

as

the

ductility.

He=-%By taking the element height h into

to account we can convert

(3.65)

the vertical

displacements

strains and the rotation into

curvature.

For

an

element

we

then obtain

(3.66)

and

conversely we may define the curvature ductility as

^=-f
whereby the equation 3.66 may also be obtained as

(3.67)

l^
ductility according
to

(3.68)

Curvature

expression 3.67

is

implemented

in the Macro model

described in sections 3.2-3.6.

for each increment and
In
an

Ductility for left and right curvature is momtored separately

absolute value of the

demands
as

largest ductility reached so far is

stored.

chapter six, ductility

described above, obtained from nonlinear time

the

history analysis, is discussed in relation

to

capacity design method.

65

CHAPTER FOUR

MICRO MODEL

4.1

Introduction

Since structural walls

mainly

act

in their

own

plane, they lend themselves ready reinforcing steel,

use

to

the

application of the plane

for separate
the their

stress relation

implemented in
as

membrane elements. This allows

as

modelling

of the constituents such that

a

concrete,

well

as

interaction,

so

composite
more more

element is formed. The

of continuum

elements makes micro models mueh

versatile than elaborate.

a macro

model, although the

separate

treatment

of the constituents is

The nonlinear

global mainly

behaviour of reinforced

concrete

structural units such

as

structural walls is

due to the tense behaviour of the concrete, and the

yielding

of

the reinforcement. A

simple

and reliable numerical model should therefore first of all

two

attempt to reasonably simulate these

As mentioned in

phenomena.
among micro models two

chapter two,

there

are

major approaches for

approach,
of

modelling

the

cracking

of concrete; the discrete

approach,
the

and the smeared

which the latter has proven to be

structures.

promising
is based

for the

modelling

of reinforced concrete that

no

The smeared

approach
element

on

assumption

single physical
an

crack is simulated
area

by

the numerical model, but the overall

an or

cracking

within

area, e.g.
as

an

associated with

with

an

integration point,

is simulated

fictitious

smeared cracks. The

early

smdies of the 1970's

commonly
when

used tension cut-off, which

essentially involves setting the stiffness matrix

and

to zero

reaching the tense strength,

see

balancing

the released concrete stress

on

the

right hand side,

e.g.

[Cerv70], and

[Chen82].
The

Performance

ofthe smeared crack models have

the tense

improved with the introduction

tension
use

of

tension

softening upon reaching

et

strength. The

softening approach was

by relating
the tension

pioneered by Hlerborg softening

cracks
to

al

[HMP76] who generalised its

the fracture energy. Other direction of

improvements, such
or

as

the introduction of rotating

(following

principal stress)

fixed cracks with variable shear

retention upon the formation of cracks have contributed to

making

the smeared crack

approach
structures.

competitive

and efficient tool for the

modelling

of reinforced concrete

66

CHAPTER FOUR

Reinforcement bars
modelled
as a

layer

Concrete

Figure 4.1 Composite model of reinforced concrete

Based upon the above mentioned this

element

chapter is devoted

to

the

development of a micro
employing

model,
on a

as a

complement to

the

macro

model of chapter three. The micro model is based

composite element

see

figure 4.1,

with separate treatment ofthe concrete and the interaction effects. used with any

the smeared crack

approach, the reinforcement steel,

The material models element. In this

developed
the

here could

essentially be

plane

stress

smdy,
were

isoparametric plane

stress membrane elements

from the

Software tis

[Abaq91]

used for the

implementation ofthe material models developed in

figure
are

chapter.

In the numerical

examples presented in chapter five the four noded element

see

CPS4 with 2x2 integration points is used,

4.2a. Other usable elements

in

including

the

eight noded CPS8

seen

in

figure 4.2b

briefly introduced
to

Appendix

D.

The numerical

integration is

carried out

according

Standard

integration procedures

and the element stiffness matrix k takes the form

JBTDBdV
strain-displacement matrix,

(4.1)

whereD is the material modulus matrix, B is the element volume,

were

and Vis the

see

e.g.

[Zien77]

or

[Bath82], The procedures for element integration

provided by
see

the Software, and will not be discussed in any further detail in this and

study,

[Abaq89]

[Abaq91].
the

We wl rather concentrate

on

development of the material model expressed by the

was

incremental modulus matrix D which

as

coded

by the user material Option,

and is

given

foUows

MICRO MODEL

67

a) Four noded CPS4

b) Eight noded CPS8

Figure 4.2 Isoparametric membrane elements [Abaq91]

da = Dde

(4.2)

where d denotes incremental stress and strain. The different constituents contribute to the

total modulus stiffness D

D
**

as

follows:

D +D +D *"c **s *^ia

(4.3)
concrete,

where

Dc represents

Ds

represents steel, and

finally Du represents
physical phenomena

the inter

action effects. The behaviour of materials and different

relevant to the overall behaviour
are

which

are

subsequeny treated in the following sections.

4.2 Material Behaviour of Concrete

This section is devoted to the smeared fictitious crack is to focus

on

development

of

useful concrete model, based upon the

approach

discussed above. Since the main

we

objective ofthe report

of the concrete
on

the

global

behaviour of walls,

limit the

modelling

the
as

micro level to

some

of the most

important phenomena

needed for this purpose, such

the behaviour upon crack formation, the so-called tension

softening,

ie

unloading

behaviour and effects associated with the cracks.

For the

compressive behaviour of concrete many material models have been proposed, mainly
be divided into linear and nonlinear elastic models, and models based

which may

68

CHAPTER FOUR

on

plasticity.

An review of such models is

given

e.g. in

[Chen82]. The behaviour in

the

compressive region
structural wall
and there is

is here assumed to be elastic, which may be if

acceptable

for many
stresses,

designs,

during

an

analysis

control is

kept of the compressive

and that in

reassurance

that

they ideally
is

do not exceed around 70 % of the

no stress

compressive
excess

strength,
the

to

which level

linearity
oceurs.

commonly observed,

of

compressive strength
the

It is also known that local

compressive nonlinearity

mostiy affects
In the

global responses in an insignificant way.

we

following presentation

will

now

develop

the concrete contribution to the

as

modulus matrix. For elastic uncracked concrete, the material is treated

and the incremental

fully isotropic,

equilibrium equation (4.2)

is

given by

da.

1
c

dE.

doA=
der. ~y)

1-VJ

0
1-v

dy

>

(4.4)

~2~. dexy

where

Ec represents Young's
concrete.

modulus for elastic small strain concrete, and

concrete

is the

Poison's ratio for

before

Although
strength,

displays
well.

small

noniinearity

in tension

reaching

the tense

the relation

(4.4) is commonly used in tension until

as

cracking

oceurs, and wl be done so in mis

study

Upon reaching
Solution to this
assumed that upon

certain tense
was
a

strength

the relation

(4.4)

must be

changed.

simple
It
was

problem

presented

in the

early study by Cervenka [Cerv70].

to a

reaching
a

principal stress corresponding

given crack strength, the

tensile stress direction.

material would crack in This the

direction
to

perpendicular to

the

principal

cracking

was

assumed

be smeared out in several

no

parallel layers,

and furthermore

assumption

was

made that

shear stress could be transferred between these

layers.

The

physical interpretation

of this model is shown in

figure 4.3.
direction

The cracked formulation of

(4.4) for cracks

in

one

was

given by ie

foUowing expression [Cerv70]

D
=

EJCC

(4.5)

MICRO MODEL

69

T=0

Figure 4.3 Cracked concrete without shear transfer across crack planes [Cerv70]
where

Ec
as

is the uncracked

Young's

modulus of concrete and C is

transformation matrix

written

COS

0 (4.6)

sin20
cos

0 sin 0

where

<f>

denotes the

angle

between the

global

x-axis and the crack

plane,

as seen

in

figure 4.3.
results for

This formulation has been used in many studies, and has

some

provided reasonable perpendicular to

the

simple problems.
was

For cracks in

second direction,

first the modulus matrix D

usually simply

set

equal

to zero

[Chen82].

simplification
planes.

of relation (4.5) and

(4.6) is the lack of shear transfer between the

crack

This fact may pose numerical

problems

due to

lack of shear stiffness. Li

simple examples of monotonicaUy

distributed reinforcement this

loaded structures with uniform geometry and

was

evenly

deficiency

however

usuaUy remedied by supporting

modelled reinforcement The
in this

neighbouring

elements and the stiffness of

orthogonally

issue of shear transfer between crack

planes wl be treated subsequeny

chapter.
(4.3)

Another
was

simplification

is that the transfer from the uncracked

to

State as

in relation

commonly performed direcy

have be balanced

the cracked State

as

in

(4.4) by cutting off the tense

from the concrete

stress, so-called "tension cut-off. The released forces

stresses
was to

originating

to

using

some

procedure.

simple commonly used method

apply these forces as externa! loads in the subsequent increment

70

CHAPTER FOUR

Figure 4.4 Typical tensile softening behaviour for unreinforced concrete

ExperimentaUy, however,
the concrete

it has been shown that upon

reaching

certain

peak stress, by
among

displays

softening behaviour, which has been carefuUy

et

studied

others Reinhardt [Rein84], and Cornelissen

al

[CHR86]. Figure 4.4 shows the typical

softening behaviour observed

behaviour is
process

for unreinforced concrete in tension. This of micro cracks within

a

softening

explained by the formation

zone, referred to as the

zone.

Various behaviour.

simplified

models have been

suggested

to

numericaUy

simulate this
more

softening
elaborate

Apart from the commonly used

were

linear

or

bilinear models,

multilinear models
also

proposed by Gustavsson [Gust85].

softening model.

Rule and Rowlands

[RR92]

proposed

nonlinear

To simulate the tensile behaviour,

we use a

model here, which

assumes

linear elastic
as

behaviour until the tense

seen

strength

is reached, followed

by

linear

softening branch,

in the

figure

4.5. When the

largest principal

stress

reaches the tense

strength of

concrete, the material becomes

stresses can

orthotropic.

In the direction normal to the crack, tense

as

still be carried

by the concrete, referred to

these cracks

the tension
a

softening effect The descending slope of softening

physical interpretation of this effect is

in which normal stresses
across

that micro cracks form in

can

so-called process zone,

be transferred. The

the stress-strain relation in the factor kts,

see

softening region is determined by

the tension

figure

4.5.

MICRO MODEL

71

'""

*'

;#S
'*

L+Le

Figure 4.5
the

Linear tension

softening model for concrete assuming smeared cracks over

length L
we use a
a

Here
to

tense

strength ft

calculated

as

Eccr, where ^r is the cracking strain set

as

0.0001,

value

commonly observed
on

in tests. This value,

e.g.

weU

as

Ec, may

to some
not

degree be dependent
treated here.

the strain rate,

see

[CB89], [Chap87], which however is

The

softening
curve

factor may be related to the fracture energy normal stress have

as a

Gf
an

as

foows. The

area

under the

representing
we

function of crack width represents the

fracture energy. Since

want to
area

always

an area

associated with

integration point,
to

we

make

sure

that for this area, the fracture energy would

correspond

the

same

in

real structure.

According

to

[CPE90] it is then possible to normalise the tension

as

softening

factor with respeet to the mesh

follows

-kc

Ec{

LJ
a

(4.7)

where L represents

length associated with the integration point,

and X is

given by

>.-**&f.
where
zone

(4.8)

Gf

is the fractore energy,

a

representing
see

the

area

under the stress

versus

process

width relation for

fictitious crack,

figure 4.6,

assumed to have behaviour

corresponding to

observed

experimental data.

72

CHAPTER FOUR

.iS'V&H

gc^j$s^wftg

if

Figure 4.6

Fracture energy

as area

under stress

versus

process zone width

[CPE90J

Gf

jcrdw
L is here assumed to be
one

(4.9)

The

length

of the sides of
areas

rectangular integration shapes,

an

area

which is close to

quadratic.

For other

integration

and element

equivalent

length may be calculated as shown by Dahlblom

The tense

and Ottosen in

[DO90]

strength

as

discussed above is further influenced

data has been

by the
on

transverse stress,

and mueh weU known

experimental

presented

the biaxial concrete

behaviour,

see

e.g.

[Kupf69].
involved

Numerically, many relatively

lished,
some

bell-shaped faure functions have been estab

see

of the most well-known

a

by Ottosen,
with the

e.g. review in

[Chen82]. These
and how it is

numerical models deal to

large degree major

compressive behaviour, global

affected

by the
we

biaxial influence. contribution to the and since

we

Since

know that the

nonlinear behaviour arises

from the local tensile

way that

phenomena,
severe

deal with structures

designed

in such

crushing

or

other

compressive nonlinearities will usually not play a major

our

role,

we are

able to

essentially

focus

interest

on

the tension/tension

region

and to

some extent

consider the effect in the

tension/compression regions. performed

for concrete,
see

Several biaxial failure tests have been

e.g.

Kupfer

[Kupf69].

We will

use a

simplified cracking

criterion shown in

figure 4.7, together with

occur

Kupfer's test

For each

principal stress direction cracking

is assumed to

when die

MICRO MODEL

73

Experimental failure criterion, according to Kupfer [Kupf69] Simplified cracking

criterion used for
numerical model

Figure 4.7 Experimental biaxial failure criterion by Kupfer [Kupf69],

and simplified cracking criterion for numerical model

tensile

strength

is reached. A

linearly decreasing cracking criterion

obtain the

in the tension/

compression zones was also

tested but did not

we

change the global behaviour. equilibrium equation for cracked concrete

(4.10)

In the local coordinate system,

da=D,d

written in matrix form

da'

'-KK
.=

0"
0

dEx de2
(4.11)

da2 dan
where

0
0

Ec
0

G. den

is the shear retention

D,

is the modulus matrix in the local coordinate system,

terms

factor, discussed subsequeny. The coupling

some

Dn

D2l Dn

have been
and

suggested by

authors

as a

small value

involving
is not

the square root of

Dn.

However since
zero.

the

physical meaning of this value

follows

fuy

clear

we set

these terms

equal to global

The

local modulus matrix in

equation (4.10)

may be transformed to the

coordinate

system

as

Dg=QD.QT

(4.12)

74

CHAPTER FOUR

where

Dg

is the modulus matrix in the

global coordinate system,

taken
as

and

is

a trans

formation matrix, shown in (4.13), axis and the normal to the crack

containing 0

the

angle

between the

global

x-

plane.
-sin 20

COS20

sin20
cos20
-sin 20

V2
sin 20

sin20
sin 20

V2
cos

(4.13)

L V2

V2

20

When E\ exceeds the value eq,

more

see

figure 4.5,

the tension

softening

does

not act

any

and the cracks have

opened

up (zero normal stress). For this

case we

obtain the

relation

0"
0

dex
d2

da,
da.
12

0 0

Ec
0

(4.14)

G. .dn.
compared
to

Unloading behaviour
uniaxial

has not been treated very mueh in the literature,

a

monotonic behaviour. There exist however,

number of

experimental
such

results from
stresses.
out

loading of concrete specimens

with

alternating
that the
seen

tensile and
a

compressive
a

Figure

4.8a shows the

stress-displacement
[RC80]. It is

behaviour of

test

carried
a

by

Reinhardt and Cornelissen

different

seen

unloading foows
that

somewhat
stiffness

path

than the

reloading path. It is also

high compressive

during unloading,

i.e. the closure of cracks, is reached further out in the tensile for

displacement region
for eycles with

eycles which
place

have reached
at a

large

maximum

displacements.

This

crack closure also tends to take

compressive
In

stress
a

which appears to increase

large maximum displacements.

The skeleton

figure 4.8b

proposed numerical model

by

Curbach

[Curb87]

is shown, which attempts to simulate the behaviour from the test

curve

shown in

figure 4.8a.
stress at

is modeed weU in

figure 4.8b

as

weU

as

the

compressive
modelled.

crack closure, but the residual tense strain at crack closure is not this

Admittedly,

proposal

has the

advantage

of

exhibiting fully isotropic

behaviour between the

eycles.
for numerical models of cyclic behaviour may be found in the

Some other

proposals

literature. Often, linear

unloading

is assumed from the

is considered

softening

branch back

to

the pro-

origin,

upon which the material

again

fully isotropic. GyUtoft [GyU83]

MICRO MODEL

75

IN/tam'.

O.N/mm'1
.A

2.0

1.0

4
/J/7K
^"
..l:

2.0

1.0

K
^

US
-1.0

-lOl

Wl

-zo

-2.0

3.01
C

-10 23
40

63

BI0

V10

12

20

40

60

80

100

120

l/jml

aj Experimental result [RC80J

&) Suggested model [Curb87]

Figure 4.8 Cyclic uniaxial behaviour of concrete posed

multilinear model for
to

unloading

behaviour.

Relatively complex
are

focal

point

models, attempting

closely

follow

experimental data,

further evaluated in

[GyU89],
residual

We here propose

simple model for cyclic behaviour which takes into

given
in

account

strain

at

crack closure. This model is

figure

4.9. It consists of linear branches

only,

for

simplieity,

and in this way is the tension

simpler

than e.g. Curbach's model.

oceurs
=

Unloading from point

to around

softening region

linearly
and
a

in

direction towards

where the cracks close,

20 percent of the
to

at a stress

level -acc

-kccft,

residual strain EcC equal the

peak

strain p, reached before the

to

unloading. By setting
data
was

factor kcc

0.5, reasonable correspondence

stress

experimental

achieved. The in the test

at

Variation of the
result from

level at crack closure appears to be mueh smaller,

than the Variation of the residual

as seen

figure 4.8a,

displacement (strain)

crack

closure. Therefore, it is believed that the

simple

model

presented

here is reasonable in this

regard.
The fact that
we

do not allow for crack closure

may

already
to

at zero stress

level,
and

as

is often

assumed in

theory,

physically

be

explained

be due to

crushing

spaUing during

of

aggregate
closure

on

the crack faces

causing gradual build-up unloading

of

compressive

stress

the

procedure.
oceurs

After crack closure further

elastically. Reloading
the energy

foows the

same

path

back

to

p. Thereby,

we

deliberately neglect

dissipation taking place

unloading
growing cracks and

during cyclic

action between crack closure and the strain from which the
we

occurred, since in earthquake-related problems

mainly

deal with

76

CHAPTER FOUR

Figure 4.9 Model of unloading behaviour in

eycles beyond the concrete tense capacity.

tension for concrete

Unloading
Once the
will

from open cracks, i.e. from strains

=

larger

than Eo,

oceurs
as

with

zero stress

until firj, from where it continues to cc

0.2eo, and then elastically

an

described above.
no

opened
in that

crack State has been reached for

integration point

tensile stress

occur

region during subsequent eycles. principal direction exceed the tense strength, a tension

Should the stress in the second

softening region will also develop in this direction:

da'
"0
'

0*
0

dx
d2
(4.15)

da2 dan

-KEC
0

G. dn
are

Due to shear retention there

still shear stresses

acting along principal

the first

opened crack

direction, and this may

direction

cause a

deviation of the second

Due to this fact,

tensile stress from the

was

orthogonal

to

the first

one.

non-orthogonal crack modelling

using non-orthogonal
some

proposed by
modelling

deBorst and Nauta

[deBN85].
in the

An alternative to

crack

is to aow for

cracking

orthogonal

direction and to i.e.

degree

correct

its effect with respeet to the shear stresses

acting,

begin

the

cracking

earlier than would be to

would be ie

case

based

only

upon

orthogonal

stress. A

further

possibility

MICRO MODEL

77

keep

orthogonal system,
was

but to let it rotate

continuously

with the current

principal
does

directions

shown

by

Rots

[Rots88]. Although a progressive crack development

change its direction at the crack tip, the concept of continuously rotating the entire existing
cracks
seems

problematic

from

physical point

of view,

although

it has sometimes
here

proven to
will be

yield good agreement

with

experimental

data. The

modelling presented

kept

within the framework of fixed crack directions, and attempt to

realistically

model the shear effects within these fixed cracks.

4.3

Aggregate

Interlock

In cracks

undergoing opening,
to as

as

well

as

in

opened cracks,

there is

some

shear trans

ferred, commonly referred

"aggregate interlock". Physically, this effect consists of

uneven

several contributing parts; cohesion, friction between the misfit due to offset surfaces. Dowel action

crack surfaces, and

arising from reinforcement bars crossing the

separately under the section
with aggregate

crack is sometimes included in this context but is treated here

"Interaction Effects". General discussions of the

interlock others.
are

phenomena associated
and Walraven

presented

in work

by Pruijssers [Prui88]

rWalr80] among

Numerically, by reducing
factor. In the
between

the combined effects of aggregate interlock, have been taken into account

the shear modulus G

by

an amount

referred

to as

the shear retention

early

finite element analyses this factor

as

was set to an

arbitrary constant value

a

zero

and one, sometimes 0.5

in

[ZP074], which often gave

set

somewhat too shear transfer

stiff Solution with

across

higher values. Attempts to

in

equal

to zero,

i.e.

no

cracks,

as

[CG71] provided reasonable solutions, but may pose numerical

difficulties in the nonlinear rnge.

retention factor

Glemberg

and Samuelsson

[GS83] proposed

shear

proportional
so as to

to

the ratio of the

a

softening

modulus and the uncracked

the

Young's modulus,
Kolmar retention. A

obtain
a

moderating effect
of different

on

equilibrium equations.
representing
the shear

[Kolm85] presents

variety

proposals
across

for

physical interpretation
with

of shear transfer

cracks is shown in

figure

4.10.

Generally,
with

low shear retention factor of around 0.01 to 0.05,

be obtained. A better Solution is to let

good agreement
for the shear

experimental data may

vary with the crack

opening strain,
retention factor upon

i.e. Mode I behaviour. Rots

as a

[Rots84]

derived

expressions

function of the

opening strain. The first of these expressions is based

and Loeber

experimental

work

by Paulay

[PL84], and takes the form

78

CHAPTER FOUR

yG

Figure 4.10
1
=

Shear transfer across cracks

l + 4447e.

(4.16)

The second
as

expression is

based upon work

by

Bazant and Gambarova

[BG80] and reads

foows

1
=

4762

1346^
curves

(4.17)

where is the strain in the direction normal to the crack The and

of equations (4.16)

(4.17)

are

shown in

figure 4.11.
experimental study
on a

Some interest has been directed towards

and Mode n, i.e.

mixed mode

(Mode I

opening mode

and

shearing mode) behaviour,

degree

see

e.g. Hassanzadeh

[Hass91], [Hass92],
Mohamed

Bazant and Gambarova

It has been observed that

[BG84], Pruijssers [Prui88], and Noorusome

[Noor92].

of path

dependency

between

the two modes exists, but it is

generally concluded
taking into
was

that further research is necessary. An

expression
ratio I y

for the shear retention

as
=

account the relation between normal and

shear strain

well

as

the aggregate size

suggested Pruijssers [Prui85],

reads

where the

3 gave

a curve

similar to the formula

by [Rots84]. Higher ratios gave

lower

retention for

given

normal strain. The

expression

with
P

1
=

PE.+1

(4.18)

2500
=

0.14

0.76-0.16-^- 1-exp
7

H>^

\Y\

(4.19)

'"/

MICRO MODEL

79

10

3
i

1
1

Rots

0.8

1
1
1 i

0.6

1
0.4

tt
\\

0.2

\\ \
>

^^ZZZmmT
i
4

<

10

10"3

Figure 4.11 Shear retention factor , [Rots84], [BG84]

1.0

Ornax=l9,T m
I

0.9

Enr

Y
1

0.8 0.7
0.6 0.5 O.C 0.3

3
in

li
II

ll

0.2
Q1

k
>.

Vv
0
1

^r= rrr 23456789

10
-3

Enn-10

Figure 4.12 Shear retention factor , [Prui85]

where

Dmax

is the diameter of the

largest aggregate.
to

For

some constant

ratios of

7%

the

expression
Here

is

plotted

for

Dma equal

19 mm, in

figure

4.12.

we

mainly
on

use a

pure Mode I model, and choose the

expression derived by Rots

simplieity. However,
behaviour wiU be
the

[Rots84], based

the work

by Paulay

and Loeber, due to its here

on

influence of the other discussed in

proposals presented

the

global

briefly

chapter

five.

80

CHAPTER FOUR

It appears clear that

during cyclic behaviour,

such

as

is the

case

during earthquake

action, the shear retention factor should gradually decrease with the number of eycles,
due to

grinding effects.
was

It

is, however, unclear how this decrease would affect the relation

4.16, which

established for monotonic behaviour. Furthermore, the

unloading
path

shear

modulus has been observed from

experiments
one

to

be

higher
but

than the

loading

modulus. back
as

This would
indicated take

imply
the

that for

unloading,
seen

should not follow the

a

same curve

by

curve

by Rots,

in

figure 4.11,

different

which would
in his
to the

some convex

shape.

This effect

was

taken into account

by Skrikerud [Skri82]

model for element

sptting, whose results, however,

may not be

direcy applicable
an area

problem
more

of the smeared crack formulation for reinforced concrete. This is

where

research is needed.
use

We

equation

4.16 therefore in

an

unaltered form for monotonic

as

we

as

for

cyclic
we

behaviour. For the closed crack branch with elastic almost fll shear retention,
or

compressive stiffness, however

undamaged
the crack faces,

employ here

rather,

factor of 90 % of the
on

isotropic drop

value. The

slight

reduetion allows for

more

grinding damage

oceurring during cyclic

a

behaviour. A

refined Solution would suggest this value to

few percent at each

cycle.

4.4 Material Behaviour of Reinforcement Steel

The reinforcement bars may be modelled

discrete model,
or

according
see

to two

basic types, either

by

the the

by

the smeared model,

figure

4.13. For discrete

modeUing

mesh size may be made to

may however often

correspond

to

the distance between the

reinforcing bars, which

number of equivalent

give
an

very elaborate mesh models. The other alternative is to model the

reinforcement within

element

lumped

in such

a manner

that

reinforcing
are

bars

are

connected between

adjacent nodes.
assembled

In either case the

reinforcing bars
early

modelled

as truss

elements which

are

separately from the concrete element reinforcing

was

as

shown in

figure 4.13b.

The discrete

modelling

of

used mueh in

studies of nonlinear concrete models.

The extension of

lumping

reinforcement within

an

element will result in the smeared

reinforcement model, shown in continuous

figure 4.13c,

in which the
as

reinforcing bars

create a

layer for each direction of bars,

and enter

uniaxial contributions into the

material modulus matrix.

Smeared

modelling is

more

convenient from

user's

point of view, since mainly efficieny

the

reinforcement ratio and steel

properties

of each direction need to be entered. The

treated
more

additional fact that interaction effects

maybe

with the smeared

MICRO MODEL

81

Reinforcement

lumped
v

into truss elements

-I

a) Prototype reinforcement

b) Discrete reinforcement model

Vertical reinforce
ment

layer.

Concrete

Horizontal reinforce
ment

layer

c) Smeared reinforcement model

Figure 4.13

Basic types

of reinforcement models
reinforcement model. Since

modelling approach, gives

we

us reason to

implement a smeared vertically,

assume

bars

only horizontally

and

the transformation to the

global

coordinate system is omitted. The smeared model has

with

proved

to

give good agreement

evenly
distributed

experimental data,
large

as

long

as

the reinforcement bars have been The

and the wa geometry has not

with

changed suddenly [Chen82].

by

occurrence

of few bars

diameters has been shown to be better modelled

a

the discrete method.

that

Drr

[Drr80] suggested

combination of both methods

arguing

uniformly

rein

forced parts should be modeed

bars should be modelled

by

the smeared method, and in addition,

single heavier
some

by

the discrete method which would also allow for

82

CHAPTER FOUR

nonlinear bond

slip Springs

to

be

applied,

as

studied for was in [Drr80] and for rein

forced concrete frames

by Glemberg [Glem84]. mainly

focus the

Since in this

study

we

on

global behaviour
vertically
and

of well-detailed and

uniformly reinforced walls, the smeared reinforcement model is used for the modelling
all reinforcement. Different reinforcement ratios

of

horizontay however are

same

generay applied,
mesh.

and the ratio may be altered between different elements in the

The reinforcement contribution

Ds

to

the modulus matrix is, for bars in horizontal and

vertical direction

only,

obtained

as

PA D.=
o

PA
0

(4.20)

where p denotes the reinforcement ratio. It is noted that and the smeared reinforcement
acts as trusses

Ds
in

contains

no

coupling

terms,

in its pwn direction. The

seen

elastic-linearly

hardening

model used for the reinforcement steel is

figure 4.14.

-i

Of E

r
7
k
yC-S

A
i

Figure 4.14 Elastic-linearly hardening model for reinforcement

The
oceurs

hardening

ratio ay may be

given by

the

user.

For

cyclic behaviour unloading

with elastic stiffness and

same as

compressive yield

stress and

hardening

are

assumed to the
a

be the

for the tense

region.

In both the tensile and the

compressive regions, during yielding

at

skeleton curve, i.e. the strain

hardening branch,
to

wl be foowed

eycles,

which

more or

less

corresponds

kinematic

hardening.

This behaviour shows

MICRO MODEL

83

reasonable agreement with

experimental

data for reinforcement bars

subjected
where

to

cyclic

action,
at

as

opposed to the sometimes applied kinematic hardening concept

oceurs at a

yielding

eycles

preselected yield stress.

4.5 Interaction between Concrete and Steel

The interaction between reinforcement steel and concrete may be divided into two main

categories:

stifferiing

effect under tension of concrete between cracks

across

involving

bond

behaviour, and secondly, dowel action of reinforcement

The first category, known
as

opened cracks.

the "tension

stiffening" effect, is due to the fact that when

behaviour of this

cracking
thus
treated

oceurs, the concrete located between the cracks still acts under tense stress and

gives a stiffness contribution. The physical

early
among others also made

phenomenon was
to treat

by

Bachmann

[Bach67]. Numerical attempts

et

tension and

stiffening mainly
steel,
a

were

by Cervenka

al [CPE90],

Dinges [Ding85], Gupta

Maestrini [GM90], and

Koegger [K0II88]. Numerically,

by modifying

this effect has been simulated

curve

in three ways; either

the stress-strain
or

for the reinforcement

the effect
one as

or

by modifying description

the behaviour of the concrete,

lasy, by modelling
most
a

separate fictitious material. The last way of modelling is the

a

complex

and

involves

of the effect

along

the rebars, and then

transformation to the

cracked local coordinate system.

The tension 4.15.

stiffening

effect in the direction of

across

reinforcing bars

is shown in

figure
As

Upon

formation of cracks

the reinforcement the interaction

begins.

tension

softening

of the concrete has terminated, the interaction wl be constant at a stress

ats of about 0.4 ft until the steel starts to

yield, whereupon

the tension

stiffening

effect

will

cease to act

and ie interaction stress

drops

to zero. This interaction

effect may be
the stress-

transformed to the cracked concrete coordinate system, and wl then

strain relation of the concrete
sion
as

modify

shown in

figure 4.16,
stepwise

i.e. for any

given

strain the ten

otherwise.

stiffening

allows the concrete to carry

higher tensile stress than it would

behaviour of

In order to

simplify
shown in with
a

our

modeing
for

the

figure
a

4.16 may be

smoothed,

as

figure

4.17. Some researchers

attempted
al

nonlinear tension
et

stiffening path

long tail,

example Hayami

et

[HMM91], Ohomori

al

[OTTKW89], and Rothe and Knig [RK88].

It has been shown

experimentaUy that tension stiffening

separately

acts

mainly in

the direction of

the reinforcement, and the best way would be to add the effect
to

treat

the effect

numericaUy or

the reinforcement modulus matrix. However, it is also found that the ten-

84

CHAPTER FOUR

er

Figure 4.15

Tension

stiffening effect in direction of reinforcement

Figure 4.16
sion

Tension

stiffening transformed

to

cracked concrete coordinate system

stiffening effect mainly acts during

by

the first few

eycles during cyclic action.

a

Thereafter, when cracks have been opened and closed

carried

few times, the tense stresses,

not

the concrete between the cracks, become small and do

to the overa

contribute it is

significany

stiffness any

more.

In

earthquake engineering problems

quite important to realisticaUy model the behaviour after tension stiffening no longer acts.
For the purposes in this report it appears

adequate therefore eycles,

to

have the

Option

of

including

the effect in

limited way for the first

and it wl be

performed by

modifying

the tense behaviour of unreinforced concrete

stress

(figure 4.5) by superimposing

figure 4.16.

the additional

carrying capacity

of

figures 4.15,

as

shown in

MICRO MODEL

85

Figure 4.17 Smoothed tension stiffening

We wl

model with

linearly descending branch

stiffening by applying
linear

use a

smoothed modification for the tension

branch from the

point

where the tense

strength

is reached to the termination of the

tension

stiffening

at a the strain Eots- The strain ots

would in fact

always vary somewhat,

achieve the
mentioned
to

due to the direction of cracks in relation to reinforcement directions.

By superimposing advantage
the of
a

ie tension

stiffening effect on

the tension

softening,

we

less

abrupt

loss of stress, and

largely

avoid the

previously

considerations concerned with the fractore energy and its mesh

dependence, compared

problems

of unreinforced concrete. We material parameter, with

are now

able to treat the tension

taken

stiffening to
smoothed

some extent as a

some care

concerning the

reinforcement
a

ratios and directions.

Dinges [Ding85] conservatively suggested

on

that for

tension

stiffening effect superimposed

as

the concrete, the

stiffening would cease to act at

directions
wl

Eots taken

10^. Based

on

the

assumption

of uniform reinforcement in both directions

and reinforcement ratios which do not differ

use

considerably between the

10ecr
as an

we

ots

ranging

from

Dinges' Suggestion

of

up to about

20ecr, which is

in the

vicinity
use

of Ey for steel and thus is to be

same

regarded

upper limit. For

simplieity,
for the

we

also

the

unloading behaviour

from the tension

stiffening

zone as

softening

zone

discussed earlier.

qualitative idea of the resulting composite modulus

in the

for

an

integration point is given stiffening acting until

runs

in

figure 4.18.

The effect of the fictitious unreinforced concrete in addition to the

seen

reinforcement is
the onset of

figure as

well

as

the smoothed tension

yielding.

This

figure

is

only

correct
as

if the

reinforcing

orthogonal

to

crack planes, otherwise the tension

stiffening

implemented

in this model wl not cease

86

CHAPTER FOUR

Composite
modulus

Yielding reinforcement
Elastic reinforcement

Unreinforced concrete

Tension

stiffening
Strain

Figure 4.18 Qualitative build-up of composite modulus

exacy

at

the onset of

yielding.

It should further be noticed that the

resulting modulus at
become

the tension

stiffening phase

wl also

depend

mueh

on

ie

area

of reinforcement If the
even

reinforcement ratio is low, the

resulting

modulus wl be low and may

or

negative. It may

occur, at reinforcement ratios which are around

to

below the minimum

some

reinforcing according neighbouring

the SIA Standard 162 (0.20 % for structural was), that

element may not exhibit

positive-deflnite behaviour during a few increments. Usuay,

a

the

elements have

supporting action,
the

and numerical difficulties at tension

stiffening were generally not encountered.

The effect of tension the

stiffening

on

reinforcing steel wl also be briefly discussed in bridge

a

foUowing.

The

reinforcing

bars which

crack must carry the

stresses not

only

of the steel

area

but also the force

resulting

from the tense stresses in the uncracked

concrete sections. This causes the stress in the

reinforcing

bar to be

higher over the

a

crack

than in the uncracked section,

see

detailed discussion [Bach67]. For

smeared rein
stress to

forcing
further

model it would be necessary to aow for this effect

by adding tense

the

smeared

reinforcing bars,

and leave the

as an

yield

level at its correct State. This may also be

simplified and regarded stiffening.

increase of the

Young's modulus

for the

reinforcing

due to the tension

The modification of the reinforcement modulus is shown

qualitatively

in

figure 4.19.

Additional steel force

across

cracks due to tension

stiffening over area Ac

(4.21)

<*uK
f.

may be transformed into additional steel stress

MICRO MODEL

87

**-

a) Additional stress carried across cracks

b) Simplified modification of

Young's modulus for

Figure 4.19
Tension

steel

stiffening effect on reinforcement

/.=

L.-BA-3.
(4.22)

which after

bud up
in

phase

should be added

parallel

to the

original stress strain

relation

of steel,
in

as seen

figure 4.19a.

For numerical

modelling,

die

simplified approach, given

seen

figure 4.19b, may be adequate. Here, the effect of the tension stiffening,
an

in

figure

4.19a, is replaced by
may be

increased

Young's modulus

for the steel. The increased modulus

approximated as

*,

_/,+/
=

(4.23)

simple calculation example with

stress

concrete

tensile

strength

of 3.0 MPa, tension

a

stiffening
ratio of

taken

as

40 % of this value, i.e. 1.2 MPa, and

assuming
a

reinforcement
of

one

percent

an

original Young's

modulus of 200000 MPa and

yield strength

500 MPa

gives the foowing modification.

J*

=-3s.
p

120 MPa

0.01

=fy=
y

500
=

0.0025

E.

200000

88

CHAPTER FOUR

500 + 100
g

SM

240000 MPa

0.0025

From

expression (4.22)
the cracks

it is

seen

that the effect becomes

larger

for smaller

reinforcement ratios. This is

carried
across

physicay
means

feasible in
a

so

far

as

the tense stresses must be

area.

by

of

smaller reinforcement
concrete

At smaller bar

diameters the bond relation between bar and

may influence the

simplification

described here.

In

earthquake engineering problems involving cyclic behaviour cracks

open up and

close

repeatedly,

and after

few eycles

essentiaUy no tense
effect
as an no

stresses are carried

oceurs.

through

the concrete, and the tension

stiffening
given

longer

Since the

Young's

modulus ofthe reinforcement may be

material model

input parameter ofthe implemented user

(see Appendix B) nothing in particular is prescribed in this question. This

for the best

is rather

problem-related question:
experiment
an

possible numerical Simulation of

corrected
be left
to some

monotonic

the modulus should

probably be preferably
advanced

degree.
to better

However, for
model ie

earthquake
with

Simulation it should
concrete. An

unchanged

eycles

damaged

Option

would be to include the and reset its value to the

possibility of modified Young's modulus for a few initial eycles

unmodified level.
We wl

lastiy

turn

briefly

to

the second interaction effect mentioned. Across

cause

opened

cracks the tensile reinforcement wl

dowel action, which may be modelled

to

numericaUy by a dowel modulus matrix, added

effect of this action is still

the concrete modulus matrix. Since the fraction

being researched

it is

currently modelled as a numerical

related
to

of the concrete modulus G, and to

some extent

the reinforcement ratio of each

direction. Although it is assumed that the dowel action also affects the modulus stiffness
terms

d\\, du, and du,

a

the entire dowel action is located at

^33

in this model. In

some

examples

numerical dowel modulus stiffness

corresponding to

between 2% and 5% G,

devered reasonable results. For the sake of

simplieity the

dowel contribution for the

micro model in this report is added into the local system, and

together

with the concrete

contribution rotated

to

the

global system.
to

Hereby,

the

description of the various contributions completed.

The micro model is

the

composite modulus of the

number of tests in the

micro model is

subjected

to a

foUowing chapter involving comparison with experimental data.

89

CHAPTER FIVE

NUMERICAL

EXAMPLES

5.1

Implementation developed
in

The models Software

chapters
a

three and four

were

implemented

in the

Abaqus
for the

[Abaq91],

which is
are

general purpose

finite element code. The

reasons

selection of this Software

discussed

by Wenk and Bachmann [WB90] and by Wenk

[Wenk90].
The
macro

model of
an

chapter

three

was

coded

by

means

of the User Element

Option.
own

With tiiis

Option

arbitrary
a

element behaviour may be described

by

the user's

Fortran code contained in Software via

a user

subroutine, which is connected

to the other

parts of the

element interface. This interface

user

essentiaUy dehvers displacements for

right
as a

each increment and the

element subroutine retorns the stiffness matrix and the

hand side (force) vector for each increment. Each of the nonlinear
user

Springs was

coded

element for initial test purposes. For

macro

wall section

as

shown in

figure 3.3,
was

the also

closed form stiffness matrix of the coded

as a user

element

developed

in section 3.7

element
was

Likewise, the micro model of chapter four

coded

by

means

of the User Material

that the material Standard
an

Option.

This

Option

is similar to the

user

element

Option, except

behaviour

only needs to be described by the user at an integration point

as

library

elements may then be used, such number of

various

plane

stress

elements with

arbitrary
user

integration points.

The interface devers the strain

increments, and the

side (stress)
are

subroutine

supplies
user

the material modulus matrix and the

user

right hand

vector.

For both the

element and

material

options

State

variables

used to stre

quantities

like accumulated stresses etc., which need to be accessed for correct

formulation in
subroutine. For

subsequent increments.

These State variables

see

are

updated

in the

user

description of the

State

variables,

[WLB93].

For

dynamic problems

the structural

mass was
a

modelled

by point

mass

elements

at

floor levels, with the

horizontal
one.

mass

acting only

in

preselected direction,

chosen here to be the

An introduction to the

use

of the models

developed in chapters three

model in

and four is

given

in the form of model in

users' manual for the

macro

Appendix A,

and for the micro

Appendix B.

90

CHAPTER FIVE

5.2

Analysis

Procedure

5.2.1 General

The used Software deals with Fixed

or a

[Abaq91] provides analysis procedures

analysis,
such
as

divided into steps. Each step

or

separate

static monotonic,

eigenvalue,

dynamic analysis.
loads followed
a

automatic time increment

a

length may be selected within each step. Several steps

as

may be executed in

sequence, such

e.g.

static

analysis for gravity

by

dynamic analysis for earthquake

are

action. The stress and strain states at the end of

preceding step
The
were

kept as initial values for the foUowing step.

At first

analyses in this report were generaUy carried out as foows:

at

gravity loads
user

applied

floor levels

during

static monotonic

analysis step. During

this step,

supplied behaviour was

load step,
structure.
a

assumed to remain within the elastic rnge.

out

FoUowing the gravity

to

nonlinear

analysis is carried

by applying

lateral action

the tested load

This nonlinear
or a

analysis may either consist of a monotonicaUy apped static

in
a

one

step,

cyclic static load apped in several steps. Furthermore, it may consist of

with
a

dynamic analysis,

ground motion

as

input carried out in one step.

5.2.2 Time

integration
procedures
foUow Standard methods described in

The Solution

[Abaq89].

For the Solution

of nonlinear

dynamic problems direct time integration methods are generaUy used,

at

which

satisfy the equations of motion

For the time

discrete times, here denoted

n.

integration

the

Hber-Hughes-Taylor
an

ct-scheme is used,

see

e.g.

[HHT77], [Hugh87], or [Abaq89] which is

extension ofthe

original Hber ct-method

is

[Hilb76]. Only

brief introduction
on

to

the

integration scheme
family of equations,

given

here. The

integration

scheme is based
as

the Newmark

which for time n+1

may be written

H&mti+Cmtl + KUltl
U^=Un + 6tH +

FlM

(5.1)

^[(l-2)n

2,.+1]

(5.2)

t/.+1=l/

Af[(l-y)lU+1]

<5-3>

NUMERICAL EXAMPLES

91

where M, C and K represent the

global

mass-,

damping-,

and stiffness matrices,

and

respectively, displacement

and

, ,

and U represent the

global acceleration-, velocity-,

vectors,

respectively. Lastly,

F represents the

externally applied force,

generaUy varying

with time.

The Solution is based

on a

predictor-corrector procedure. The predictors

are

defined

as

n+1

=UH

AtH+^-{l- 2)_

(5.4)

.+1=n+(l-y)to.
With known values for

(5.5)

displacement

and

velocity

at

time

0 the process may be

started

by obtaining the acceleration

as

M0=F-C0-KU0
or

(5.6)

speeifying it direcy. Then

,.+i is determined as
(5.7)

(M+yAfC+Mr2*)^ =F.+1-CUH+l-Kn+l
where the factors

and
=

y are chosen 0

by

the user,

see

further e.g.

[Hugh87].

From

(5.7)

it may be noted that

permits expUcit integration, (uncoupled system diagonal),

otherwise
are

of

equations
out.

provided
corrected

that C and M

are

impcit integration

is carried

The

displacements

and velocities

obtained as

UK+1

n+1+At2H+l
+

(5.8)

Vx= 6^

7*0^
some

(5.9)

The a-scheme introduces

algorithmic damping without deteriorating

as

aecuracy

by

retaining relations (5.2)

-

and

(5.3) and modifying (5.1)

foows

MH+1 + (1 cc)Cn+l aC. + (1 + cc)KUH+l

-

aKU.

F(tH+tt)

(5.10)

where

(tH+a)

tH+l

aAt. In [HHT77] the

=

following choices
and y
=

for the parameters

are

recommended:

ae[-^,0], (l-a)2/A

(l-2a)/2.

92

CHAPTER FIVE

Low values for

oc

such

as

-0.05

were

found

to

be favourable
at

according

to

among

others

[Abaq89], giving
problems

very

slight algorithmic damping

lower modes, and is ie

(changeable) default value in [Abaq91].

Since the
are

generally nonlinear,
are

the above

equations

are

solved for

incremental unknowns, which

increment

then accumulated. Fixed and

automatically changing
chosen for a

lengths

are

available in

[Abaq91],

the former

being

problems

solved in this report.

5.2.3 Residual forces

For nonlinear
an

problems,

the internal element forces, which result upon the Solution of with the

increment

are

generally not in eqratibrium

apped

external force.

Thereby,

unbalanced residual forces forces.

develop. Different procedures exist to deal with these residual

Only

short discussion of this

problem is given

here.

Purely incremental procedures [Yane82]

The event to event

do not correct for the residual forces, and Solution. solutions

to

thereby generaUy require very smaU increments in order not to deteriorate the
until
the

procedure [Lind88], [SB90] essentiaUy performs linear

This is in

nonlinearity oceurs.

practice performed by scaling the

a

Solution

reach

nontinearity, and is best suited for problems with

The incremental-iterative procedures

few

sharp stiffness changes. The

effective for

procedure is competitive only for relatively smaU Systems of equations. [Abaq89]

are

the

most

arbitrary

and

large Systems. They may be divided into fuU and modified Newton-Raphson procedures.
Whereas the former

generally update

the internal forces

by recalculating
on

the element

matrices, the latter adjust the residuals by applying equibrating forces

nodes. For the used.

the unbalanced

problems computed
tolerance values

in this report the fll

were

Newton-Raphson procedure
one

was

Typical

selected in the rnge of

are

percent of maximum

element forces. Within each increment iterations

performed (update of stiffness matrix

are met

during iterations) until

each increment, incremental

die

preselected tolerances

Based upon the stiffness

at

(modulus) matrix obtained from user elements during

the
are

assembling

the

beginning of
the obtained

predicted

element forces

(stresses)
are

computed, using
as

displacements. Residual forces

computed
the

the difference of the

predicted and calculated

tolerances

element forces at the end of the increment. If the residual forces

are

exceed the tolerances, iterations

are met

performed, updating

predicted values,

until the

NUMERICAL EXAMPLES

93

5.2.4

Damping
for the

For

dynamic analysis the hysteretic behaviour accounts algorithmic damping provide

some

major energy dissipation.

rnge of interest it
was

Since the

was

very

sght in the frequency damping.

by
The

decided to
viscous

additional viscous

major reason for this is

that

damping may be better controUed physicaUy than the algorithmic damping. damping
is

This additional

provided

for

the introduction of

Rayleigh damping

matrix C,

consisting of one mass proportional part and one stiffness proportional part.
(5.11)

a + a2M
mass

The

and stiffness coeffieients

were

determined in the weU-known

see

manner

by

selecting
lowest

desired

damping

ratio at two

frequencies,

e.g.

[Bath82], whereby the

frequency

is chosen well below the elastic first mode, to account for the due to

decreasing eigenfrequency frequency is important.

For the taken from

yielding during

the time

history analysis. The second

an

upper mode above which the viscous

damping

will be

examples with

die

eight-storey structural waU buding, presented later in this

chapter,

around two percent viscous

at

damping was

selected at around half the

as

frequency
0.12.

ofthe elastic first mode and

the second mode,

giving ax

0.0009 and a2

as

5.2.5 Ground motion

The Software

generaUy permits ground

In the

motion to be

apped

in three

perpendicular

directions into

simultaneously.

examples studied
motion in used

in this report, which

were

simplified
Time

planar analysis,

horizontal

ground

one

direction only

was

apped.

history ground accelerations

elastic

were

as

input.
the

One of these consists of an

artificiaUy generated ground acceleration, compatible to

design

response spectrum of the Swiss Standard SIA 160

[SIA160] for five

percent damping and medium stiff so.

The other

ground

acceleration consists of the recorded N-S component of the 1976

at

Friu

earthquake, recorded ground

Tolmezzo.
are

Both

accelerations

shown in section 5.4.2 and

are

discussed further in

connection with the numerical examples.

94

CHAPTER FIVE

5.3 Selection of Numerical

Examples

Due to the nonbnear behaviour of structural walls, the

only method to check the reUabity

data. Therefore in
we

of

numerical model is

by comparison
shear wall

with

experimental

carry out

some

elementary comparisons between

data from
a

the models

developed

chapters

three and four

and

experimental
is

specimen.

The selection and

description of this
models in

specimen

presented below

in section 5.3.1.

In addition to the

reliabity

check it is desirable to

use

the

developed

Performance check on
check is

the behaviour of a

capacity designed multi-storey buding. Such a

and the

performed in this chapter as well,

buding is presented in

section 5.3.2.

5.3.1 Test

specimen
structural waU

large
A

number of tests

on

specimens of different scale have been carried

It is, however, difficult to find tests
on

out.

good review is given by Abrams [Abra91].

as a

carried out

realistic Simulation of

earthquake action
to

taU slender walls. It is in order to avoid scale

desirable

to use a test
as

specimen

as

close

fuU scale

as

possible

effects such

the wrong amount of concrete

concrete

cracking, difficulty in reproducing

For

reinforcing and
the
A

aggregate

properly,

etc.

dynamic

tests the scale effects on

frequency is an added factor.

large part of the tests
concern

with full-scale

or

close to fu scale i.e. walls with

specimens, documented in
an

the terature,

"squat shear walls",

aspect ratio (height

are

to

length)
as

smaller than two.

(In [PBM90] was with aspect ratios below three

regarded

squat walls).

In several other tests the the lower stories, and

stories
as

specimen

was

assumed to be part of

taU wall,

representing

subjected to a high vertical force simulating gravity load from Upper

shear force

weU

as a

large

simulating the accumulated inertia forces arising from overturning

moment

the fictitious upper stories. However, the which may be of

a

from the upper stories,

not

considerable

magnitude for multi-storey buildings, is usuay

included in the test

Among tests with specimens

an

assumed to be part of a wa, the test series of Vaenas,

Bertero, and Popov [ VBP79] included an attempt to simulate the overturning moment as
action
on

the

specimen.

This test series also

provided extensive

test data

which lends

itself to

comparisons with the numerical

models

developed in this report

NUMERICAL EXAMPLES

95

76mm-

2.388m-

1.680m-

-1016m

1016m

406 76mm SLAB

0.30Smt

|
2 AT 76

mr

T^ -r.

ml

.-.

...

.-..-,
ui
i

_-.-|ii

m.t

nr
0914 IS mm
m

l 0.152

I 1SLA8

76 rrfrt

mm

0.102

WALL

_i
4.:S2m 0.914 m
.

EACH WAY AND EACH FACE

76mmSLA3
WALL 0.O2m THICK
r-,i

76

mm

SLAB

T
76mm l.2!9m

-0254m SO. COUJMN WITH 8*6

3 6

0.254

SO COLUMN

S-r1

76mm
G.H. 7 HOOPS

Llj

4>

mm

0.356m

CLEAR BOTH SlOES

95 mm

ZTmTSrl

Tmim.m.m.m.rmm.m.
C8m
1.1.
i. it

it-<<-

V
F00TW3

jl
I! :i

ii

ii 'i
ii

ii
"

jLjJ

ii ii M aUu.JL.
'rr
--

'!

|l
ll
11

F00T1NG

LJ."

at._

dti

1_ .JL
-0.660m
95 mm

3.098

a) Elevation

b) Vertical cross section

0.254 u*.
m

LstfS

r-t =102 rrm

A
Gft6EMD.7ATI^
L
=

2AT76rm)

nsz

2.388m

c) Horizontal wall cross section

Figure 5.1 Three-storey wall specimen [VBP79]

The
tests

comprised

the

experimental study of two three-storey

were

wall

specimens

in the

scale of 1:3. One wall had

cross

protruding boundary elements and the other was rectangular in subjected

to

section. Both specimens

gravity loads, monotonic and cyclic

lateral loads combined with monotonic and cycc vertical loads.

Some of the results from this

in

experimental study were also used for reliabity checks

as

previous studies

on

numerical wall models such

in

[VB87],

and in

[Lind89]. We

use

here the wa with

boundary elements for the purpose of checking the reliabity of the

96

CHAPTER FIVE

18 601

~T
6.1m

9 AT 2 745m

203mm

"FLAT SLA3~j
o

28 36m

24.7m

6IOmmx6lOmmCOL
o

\
7 AT 6.1m 42 7m
a

jo'ii'g
3 66m

54.9m

^508mm*5O6mm

COL

S5SSSSB

SPANDREL
BEAMS

&J Elevation

-SHEAR WALL-

0.644

V'l

J0.644V
J434
kN

434kNl
6.1m

J
6.1m

61m

6.401m

a)Plan

c) Section force application

on

model

Figure 5.2 Ten-storey building,

numerical models

with model

of three lowest stories

developed

in this report Elevation and section of the test

as a

specimen are
The section
as a

shown in

figure

5.1. The

specimen was fabricated

a

1:3 scale model ofthe three lowest

seen

storeys of the structural wall of

forces of the wa force
at

ten-storey

building,

in

figure 5.2.

the fourth floor level

to the

were

transferred to the scale model

a

shear
as a

(shear force applied

top of the model) and

bending

moment

(applied

vertical force

couple,
to

i.e. two vertical forces

the shear force
as

applied

at

the fourth floor with

opposite

sign,

and

coupled

shown in

figure 5.2).
moment

It should be noted that

although usuaUy
history analysis
not

better than

omitting the bending

entirely,

this is also

simplification ofthe relation between

the

the section forces which

assumes

that

during a time
in Table 5.1.

bending moment is proportional to the shear force, which is generaUy

the

case.

The basic mechanical

properties

ofthe test

specimen

are

given

NUMERICAL EXAMPLES

97

Concrete compressive strength

34.8 MPa
27 900 MPa 3.48 MPa

Young's modulus for small strain concrete

Concrete tense strength
Concrete tense strain Reinforcement yield strength, boundary element Reinforcement yield strength, web

0.0001
444 MPa

507 MPa 211 000 MPa about 1% of elastic modulus

Young's modulus,

all bars

Strain hardening, all bars

Table 5.1 Mechanical properties ofl:3 scale

test specimen

[VBP79]

5.3.2

Capacity designed multi-storey wall building

numerical models, after the

The
use

objective with the

reability

tests are

satisfactory,

is to

the models in the

Performance check of capacity designed buildings. We use here an

which is located in Switzerland, in seismic 3b

eight-storey building, which essentiaUy corresponds to the one used in [BWL92b].

The

eight-storey office building,

to

zone

according

the Swiss Standard

[SIA 160], is horizontally stabised by structural walls,

not

and slender

gravity

load dominated columns carry the vertical loads

carried

by

the

wa. Plan and elevation of the

The

buding
was

are

shown in

figures 5.3
to

and 5.4.

eight-storey building
we

designed according
a

the

capacity design method

[PBM90], and
numerical tests
in

here

give only

brief

presentation
More

and the data relevant to the

performed later in this chapter.

comprehensive information is given

equivalent force method,
based on

[BWL92b].
The lateral

design forces were obtained by the

static

an

elastic

design spectrum, assuming seismic

was

zone

3b and medium stiff ground

to two

according

to

[SIA 160]. Force reduetion

levels.
to

performed according

global displacement

ductility
Q equal

Overstrength was aecounted for by reducing the design forces by a factor

design definitions
a see

0.65. For basic

Appendix F.
=

The two

ductity levels, employed

as

global displacement ductity p.A.

obtain the

3, and \La.

5,

respectively, the former known

fundamental
to

restricted
to

ductility, and the latter as fll ductity. The

spectral acceleration value
die wall
was

period of Vibration

used

found

be 1.38 Hz

according to the simplified code formula [SIA160] which does not consider

wall dimensions. A Computer evaluation

involving

only

resulted in 0.74 Hz

(elastic),

and 0.67 Hz (cracked) which is

obviously somewhat on the low side.

98

CHAPTER FIVE

444A'-
i i i

v-^

VO

4444
i i

8
VO
.,.

I
i

I
i

444
VO

4444 >-
8
VO

_4+-
6.00
| 4

6.00

6.00

00
Figure 5.3 Plan of capacity designed eight-storey wall building [BWL92b]

+Jr-

IffiP
gfSfit
r...

%m*m

rr\"<rxir'"i"

>Y%

>'

iM

mm*

IJi
v////

yrr/7 7777777

>j

6.00

6.00

6.00

Figure 5.4 Elevation of capacity designed eight-storey wall building [BWL92b]

NUMERICAL EXAMPLES

99

With the
in

slight addition of girders

was

and nonstructural elements 0.90 Hz

was

obtained

as

[BWL92b] which
With
a

also used in this

mass

study

as a

realistic

design

value between the

two extremes.

storey

of 3.08 MN
as

(2.78
=

MN for

roof) the results of the static

equivalent force calculation

and

were

obtained

ME

16.1 MN

bending
and
a

moment at
=

the base

VE

0.72 MN base shear for the restricted

ductility design,
obtained

ME

9.7 MNm and

VE

0.43 MN for the fu

to

ductility design, respectively. Using

are
=

resistance factor jr
=

equal

1.2, the corresponding flexural demands

as

Mt

16.1(1.2)

19.3

MNm, for the restricted ductity design, and M,

9.7(1.2)

11.6 MNm, for the fll

ductility design.
The material

speeifications

were set

according

to the SIA Standard 162

[SIA162]

as

follows. Concrete for walls: concrete B40/30

(design strength fc

19.5 MPa), is

Reinforcing

steel: S500

(design strength fy

460 MPa). Since this

buding

analysed

by nonlinear time history analysis, the most realistic nonlinear behaviour is desirable (no
reduetion in the
the time

input ground motion applied). Therefore, overstrength is

the

aecounted for in
mean

history analysis by adjusting

strengths

to their

effective

values

as

follows:

fy
in

550 MPa,/c

30 MPa.
mm

The wall

cross

section, shown in figures 5.5 and 5.6, is rectangular and 300

levels. The

thick
was

by

6.0

length for both ductity

buckling

criterion

according [PP92]

found to be in order. The vertical reinforcement ratios

restricted

were set as

follows: For the

ductility:

0.26% distributed

uniformly

over

the

cross

section, except atthe

6D20

D10/200

Sym

''4

r-i

500

1^/2

3000

Figure

5.5 Wall

cross

section from first storey

of restricted ductility design, Ha

6D12

D10/250

Figure 5.6 Wall cross section fromfirst storey of fll ductility design, \i\,

100

CHAPTER FIVE

ends, where there is

fll

confined

zone

of 500

mm

length with
the

1.2 % reinforcement. For the

ductity design diese

waU. The

values
a

were

altered to 0.21% and 0.45 %,

over

respectively. The
and the

horizontal reinforcement has

ratio of 0.20 %

height of the plastic hinge

are as

remaining
=

resulting design strengths

moment

at

the base of the wa

foows:

MR

19.4 MNm
and

bending
=

and

VR

2.7 MN base shear for the restricted

=

ductity
of 1.2

design,
fu
we

MR

15.8 MNm

bending moment and also VK

follows:

2.7 MN base shear for the

steel

ductility design. By assuming an overstrength factor for reinforcement

obtain the
two

X0

flexural

overstrengths
=

as

Restricted

ductity:

Oow

1.2-^

19 4

1.45

16.1

Fu

ductity:

<&.

1.2^

15.8

1.95

9.7

These values

are to

be used for the shear demand calculation which,

using

dynamic

magnification factor of
)

=1.3 + = 1.3+ 30 30

1.57

where

is the number of storeys, and inserted in the formula for shear demand [PBM90]

co

<E>

V*.

gives shear

demands of
=

1.41(1.57)0.72

1.59 MN for restricted

ductility,

and

1.95(1.57)0.43
to

1.32 MN for fu

ductity, respectively. With

the concrete contribution

the shear capacity of

Vc

vcbwd

0.91(0.3)0.8(6.0)

1.31 MN

where the concrete shear stress is taken

as

0.6J-^- 0.6,f 4;15 V 0-3(6.0) yAg

=

0-91 MPa

the

remaining

steel contribution would be rather small. However, the minimum and

requirement for horizontal reinforcing ratio of 0.20 % [SIA162],

in each face

assuming bar D10

spaced

at

250 mm,

gives a steel contribution of

NUMERICAL EXAMPLES

101

V,

aj./A^^-)m^^>0.25

139 MN

which

together

with the

concrete

contribution results in

shear

capacity

of 2.7 MN,

exceeding

the demand for both

ductity

levels. Even with

are

an

assumed material

overstrength
curtailment

of 1.4 rather than 1.2, the demands

clearly exceeded. This larger

wl
see

effective shear

capacity
the

may, however, be necessary

was

as we

later.

Reinforcing

over

height of the wa

performed according to [PBM90]. Hereby the

introduction

to

the

design of the eight-storey waU is completed.

of horizontal

Due to the

building symmetry, and with assumption

ground motion
analysis

perpendicular to the buding in the direction of the was, only half the buding needs to
be

analyzed, laterally stabised by

one

structural wa. The total

mass

used in the

for the

single wall corresponds to half the buding mass.

5.4 Macro Model Results

5.4.1

Comparison

with

experimental results

In order to check the

tests are

reliabity of the macro model developed in chapter three a number of

tests are

performed. These
an exact

carried

out as a

comparison

of the

analysis results

obtained from the numerical model with the results obtained from experimental tests. It is

clear that
one

agreement

cannot

be obtained with

relatively simple model like the

ability
of

developed

here. However, attention should rather be directed towards the

the numerical model to simulate the

major kinematic phenomena of structural was

subjected to earthquake action.

In order to fadlitate the
were

clarity of the foUowing reabity

reastic behaviour of

tests, two

major principles
weU

followed. The first is the estabUshment of a clear and relatively

a

simple test example

as
as a

which still represents clear and

reasonably
tests. The

structural wa,

simple

numerical

modeing

of this

example.

This

example

wl be foowed

throughout the reabity

of the numerical tests

second major principle is that of enhancing the

at a

clarity

presented, by essentially varying one parameter only

time, and

keeping aU other parameters fixed.

102

CHAPTER FIVE

5.4.1.1 Monotonic behaviour

One of the test

specimens
two

was

subjected

to a

static monotonic test After

appUcation of
increased tl
and

gravity

loads
a

by

vertical forces of 0.434 MN at each end of the fourth floor of the

specimen,

shear force V
a

apped

at

the fourth floor

was

monotonically

faure occurred at
was was

shear force

sghtiy

above 1.0 MN. Some

unloading

reloading couple
shown in

performed.
also

In addition to the shear

force, the aforementioned vertical force

to the

apped

and increased

according

prescribed ratio checking

of 0.644 V,

as

figure

5.2c. The results from this test

are

suited to

the influence of basic

properties

of the numerical model.

As for all finite element models the chosen element mesh will affect the obtained

Solution. For the

macro

model,

we test

this effect here

by discretising

the

three-storey test

specimen described in specimen chapter three.

from the

section 5.3.1 into three different meshes.

Figure

5.7 shows the test

with its load pattern and the three meshes for the
Macro elements formulated
a as

macro

model
are

developed

in

described in section 3.7

macro

used. The first

as seen

storey is discretised into

different number of

elements for each mesh

figure.
0.434 MN

0.434 MN

0.644 V

0.644 V

A'<%tYifYt'X<
0.914
m

MeshA

K&ii

%l
&Y&:?>>< 0.914 m
*. Jri
..mmmtiim

-v>7%

(Ar
'

MeshB

mumAwhmii

<.r*Y*y&s&
1.219
,5 m

^^m**-

MeshC

Lw= 2.388
Test

specimen

Numerical discretisation

Figure 5.7 Macro model meshes

NUMERICAL EXAMPLES

103

With wall geometry and material data from section 5.3.1, the the
macro

input
set

was

prepared

for

model, described in Appendix A, resulting in

basic

of input values

kept

for all

examples and shown in table 5.2.

0.321 m2
of inertia, /
concrete,

Cross sectional area, A Cross sectional

moment

0.193 m4

Young's modulus for smaU strain

Ec

27900 MPa

Ratio of cracked to uncracked stiffness for vertical

spring, cccr

0.5

Ratio of yielded to uncracked stiffness for vertical spring, oc? Yield moment My, (selection shown in Shear force
at

0.01 3.0 MNm

0.57 MN

Appendix C)
expr.

shear
to

cracking, Vc,

acc. to

(3.49)

Ratio of cracked

uncracked shear stiffness, ccs

0.18

Table 52 Basic input values for Macro model

test

examples
the fourth

Figure
floor

5.8 shows the

experimentay obtained horizontal displacement slip

an

at

versus

base shear for monotonic load. Some bond

of vertical bars in the

foundation
horizontal

footing

of the test

specimen occurred, causing

in

additional 10 to 15 %
in mind when

displacement included

figure 5.8, which should be kept

only the

comparing with
In the

numerical results. For the numerical models

are

mesh is

changed

and a other properties

kept constant
and

foUowing numerical examples some of the parameters of the basic set are varied, only when specifically mentioned,
1.0 MN,
are

however

are

otherwise reset to the

original value.

For the

investigation of the mesh effect a monotonic shear force is apped at the fourth
to

floor, increased

according
to

to the

pattern shown in figure 5.7,

including

the

vertical forces which

was

coupled

the shear force. The numerical

appUcation

of force

performed in

increments of 0.02 MN,

giving a total
shown in

of 50 increments.

The result is the fourth floor

displacement

figure 5.9

for the three meshes. due to

The fixed end rotation which occurred

accidentaUy at the base of the test specimen

slip

of vertical bars in the foundation is not accounted for in the numerical

seen
a

modeing.

It is

that the mesh A, with the first storey discretised into

only one macro element,

the fact that the

shows

softer stiffness, which is visible

already

at

around 0.15 MN base shear, where

to

flexural

cracking

oceurs,

again visible

upon

yielding. This is due

made here do

smaer stiffnesses

(cracking and yielding) occur over a larger part of the structure than in

the meshes with finer discretization. The

comparisons

not

exceed

shear

force of 1.0 MN

or a

fourth floor displacement of 40

mm.

Exceeding these values would

not be reastic at this floor

height with a trUinear model.

104

CHAPTER FIVE

BASE SHEAR

V(KN)

DISPLACEMENT Sjtmm)

Figure 5.8

Fourth floor monotonic horizontal

displacement

versus

base shear, obtained

from experiment [VBP79J

Mesh A
Mesh B

Mesh C

0.01

0.02

0.03

0.04

Fourth floor horizontol

displacement [m]

Figure 5.9

Mesh

effect, fourth floor horizontal displacement

NUMERICAL EXAMPLES

105

ct..

0.5
0.8

cx..

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure 5.10 Influence of ratio of cracked to uncracked stiffness of vertical spring

In the

subsequent

tests, where the effect of other parameters

are

studied, mesh A is

used in order to demonstrate the other

The next basic parameters There it is
seen

respective effects more clearly.

are

to

check

obtained

by studying

the

input properties.
as

that the ratios between cracked and uncracked concrete,

to

well

as

the

ratio between Some

yielded stiffness

uncracked stiffness of the outer vertical

both ratios
on are

Springs appear.
well
as

suggested values for these

briefly
shown.

mentioned in

chapter three as
when

in

Appendix
should be

A. However, the influence

the

global behaviour,

they

are

varied,

The first parameter tested is the ratio of cracked to uncracked stiffness of the outer

vertical

Springs,
in

Gfcr. This ratio represents the relation between the two stiffnesses

Kcr and

Ke,
of

seen

figure
and

3.19.

Varying
the

this ratio from 0.2

over

0.5 to 0.8,

by using

the mesh A

figure 5.2,
It is

applying

same

monotonic force

as

for the mesh test, the result is

shown in

figure

5.10.

seen

from

figure 5.10
test

that the choice of acr

0.2 appears

to

give the best

correspondence to the

(figure 5.8)

until

yielding oceurs. However,

one must

keep

in

mind that 10 to 15 % of the horizontal

displacement values of the test are caused by bond

106

CHAPTER FIVE

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure slip

5.11

Influence of ratio ofyielded to

uncracked stiffness

of vertical spring preceding

flexural

in the foundation. Furthermore,

some

cyclic testing
making

appears to have been

the monotonic test,

(as may be

seen

in

figure 5.8)

which may have caused

some

cracking, although unintentionally,

These circumstances

thus
a

the tested cracked stiffness smaer.

uneven concrete

together

with
a

somewhat

quality
a

of the test

specimen [VBP79], suggest

that

value for (Xcr of around 0.5

(giving

global flexural

stiffness of around 70 % of elastic,

according

to

figure 3.22)

will still be

Optimum,

if the

physical test environment is

taken into account

Next, the influence of the ratio of yielded stiffness

vertical
and

to

uncracked stiffness of the outer

Springs, ay, is tested.

in

This ratio represents the relation between the stiffnesses

are

Ky

Ke

figure
0.5

3.19. Values of 0.005 ,0.01, and 0.02

tested, with the ratio Ofcr kept

constant at

again. The result is shown in figure 5.11.

It may be dent
on

seen

from both

figures 5.10

and 5.11, that the and

ductility obtained is depen

the chosen

the chosen

cracking stiffness (figure 5.10) displacement,

on

yield stiffness
in the
are

(figure 5.11).

Both the

displacement ductility, which may be direcy

and the curvature

seen

figures

for the roof level horizontal

ductility

generally

dependent
ductility
flexural

on

the choice of these two parameters.

a

Although
a

the obtained

displacement
only

varies with

factor of almost two between

0.8
to

chosen stiffness reduetion due to

cracking from

0.2, it is also

seen

that between the factors 0.8 and 0.5

asma difference in

ductity wl occur. However, for the choice of the yield stiffness

NUMERICAL EXAMPLES

107

o.

0.14 0.18
0.25
_

a.
a.

0.01

0.02

0.03

0.04

Fourth floor horizontol

displacement [m]

Figure

5.12

Influence of ratio of cracked to uncracked stiffness of horizontal spring

as

fraction of uncracked stiffness,

ranging

from 0.005

to

0.02,

we

also obtain

displacement ductility difference of about two. cracking

stiffness and

For the normal ranges of the parameters of

yield stiffness, it appears that the yield stiffness will have the most
in the first
as

important influence on the ductity. Concerning ductility ductility,

which
we we are

place interested

in the demand of curvature

a

may obtain

the

largest ductility reached during

time

history
at

analysis. This quantity will be used for the proper design of the wall cross section
location in order
to ensure

that

that it
a

can

take this amount of curvature without faure. In the wa structure, the curvature

Coming section dealing

with

capacity designed eight-storey whereby

ductity

wl be examined and discussed,

the influence ofthe

yield stiffness

also

will be shown.

In the

same manner as

for the outer vertical

Springs, the influence of the cracking

describing
was

ratio

for the horizontal

spring

wl be shown. The factor as,

the ratio ofthe cracked

shear stiffness to the uncracked shear stiffness, which varied from the recommended minimum value 0.14

discussed in section 3.6, is

over

0.18 to 0.25, and the result is

shown in
The

figure

5.12.
to

next

parameter

be varied is the location of the relative centre of rotation

as

discussed in section 3.3, and denoted

by hc. The purely elastic derivations in that section

suggested

values of this ratio of around 2ft/3 to h/2, where h is the

to

height

of the

macro

element and hc was the distance

the top of the element

108

CHAPTER FIVE

1.0 0.9 0.8 0.7

-

ii

0.6
o o -C in

in
in

0.5
0.4
_

///

V)

0.3 ~r 0.2
0.1
0

0.01

0.02

0.03

0.04

Fourth floor horizontal

displocement [m]
centre

Figure 5.13 Influence of location ofthe

We wiU here test this parameter, h-h
c
=

of relative

rotation

expressed by

the value c, which is obtained

as

i.e. the value to be the distance of the centre of the relative rotation to the bottom of the
element related to the element

height. According to the above the suggested values would

By taking yielding into
by assigning
centre

give

in the rnge of 0.33A to 0.5h.

account, which

usuay
zero.

starts

from the bottom of the element, this value would be lower, here show the influence of this ratio
result is shown in flexural

perhaps close to

We wl

it the values 0, 0.2h, and O.Ah. The

figure

5.13. When the

of relative rotation is
moment

placed low,
in

the

yielding

will start earer

provided the bending

case.

is

increasing towards the

seen

base of the wall, which is

normally the

This effect may also be

figure 5.13.

Thereby the reliability

now turns to

tests for monotonic

behaviour are concluded, and our attention

the behaviour under

cyclic loading.

5.4.1.2

Cyclic behaviour

For

similar

test

specimen
data for

to

the

one

used for the

reliability

tests

for monotonic check the

behaviour,

some test

cyclic

tests is

also available, which

we use to

reabity

of our numerical model under

cycc loading.

NUMERICAL EXAMPLES

109

'
_

OlSPLttEMOfl'

Figure 5.14

Control displacement history for wall

test

specimen [VBP79]

Figure

5.15

Fourth

floor

horizontal

displacement

versus

base shear,

experiment

[VBP79]
The

test

specimen

was

subjected
as

to

staticaUy increasing horizontal load, coupled with

in this

the vertical force

couple

shown

previously

chapter. This load

set was

then

reversed in

number of eycles with

prescribed maximum horizontal displacement at the

shown in

fourth floor. This control

displacement history is

figure

5.14.

110

CHAPTER FIVE

1.00 0.75
0.50

i~~7i
1
^

^yj
/
~

Y/^mr^T^^ / /

0.25
o

:7^
A^^
/

0
O w o

0.25
0.50 0.75
1 00
i
i

-0.04

-0.03

-0.02

-0.01

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure 5.16

Fourth floor horizontal displacement versus base shear,

cyclic

test,

analysis

2.0

1.5
1.0 0.5
0

~T~

/\
Y
Y

/
/

\
1

0)

-0.5

/
//

^^^

-1.0
-1.5

mr^*^

V
1 1 1

-2.0 -0.01

0.01

0.02

0.03

0.04

Elongation [m]

Figure 5.17 Spring force

element, cyclic
test.

versus

elongation for left vertical spring offlrst storey macro

The fourth floor shown in

displacement

versus

the base shear from the

to

experimental
for

test

is

figure

5.15. For the numerical model the response but with the number of

cyclic static loading is clarity,

mm.

shown in

figure 5.16,
of the

eycles reduced
of around 20

to two
mm

with

fourth floor maximum horizontal

displacements

and 40

Since

repeated eycles

same

maximum

displacement did not produce any signifieant stiff-

NUMERICAL EXAMPLES

111

2.0 1.5 1.0

1\

/ /
/

/
/

[MN]

p cn o

Force

p cn

/ mS^

^^

-1.0 -1.5 -2.0 -0.01

//

0.01

0.02

0.03

0.04

Elongation [m]

Figure 5.18 Spring force versus elongation for right vertical spring offlrst storey macro
element, cyclic
1.00
test.
... .

0.75 0.50 0.25

0

^
a o

c5
Lu
o

UY / IY

^/

Yl
/

-0.25
-0.50

"

-0.75
-1.00

/Y^
i

YY
i

-0.002

-0.001

0.001

0.002

Shear deformotion

[m]
spring offlrst storey
macro

Figure 5.19 Spring force

element, cyclic
test.

versus

elongation for

horizontal

ness or

strength degradation,

the results from these two

eycles

are

suitable for

com

parison of experiment and analysis.

In order to obtain
an

idea of how the outer vertical

Springs

work

during cycc

beha

viour, the spring force

same

versus

spring elongation is

shown in

figures 5.17 and 5.18 for the

cyclic

test for the outer vertical

Springs of the first storey macro element of mesh A.

112

CHAPTER FIVE

2.0

1.5
1.0
-

^
Q> o

0.5
0

-0.5

-~

-1.0
-1.5 -2.0

iii.

-0.01

0.01

0.02

0.03

0.04

Elongation [m]

Figure 5.20 Spring force

macro

versus

elongation for central vertical spring offlrst storey

element, cyclic

test.

In

figure 5.16

change of stiffness
(e.g.
seen

may be noted

on

the

reloading branch

at a

force

level of around 0.70 MN

in the upper

right part of the plot).

wa

This is due to the

change from yielding in compression of the right edge of the

(for the model: the right

elastic behaviour. This
at

spring

seen

in

figure 5.17)

to

closed cracks i.e. when

compressive

stiffness change only takes

to

place

yielding in

tension has occurred

this side

prior

yielding in compression.
We wl also need examine the behaviour of the horizontal

to

spring during the cyclic

the

test.

This

spring,

which models the shear behaviour,

employs cyclic

relatively simple chapter

bilinear three.

origin

oriented

hysteretic model, whose justification spring elongation during

5.19. the

was

discussed in

Spring
macro

force

versus

test

is shown for the first

storey

element in

figure
the

In order to

complete

study of the kinematic behaviour the spring force

versus

elongation

for the central vertical

As
seen

spring
this

of the first storey

is active

macro

element is shown in
The small

figure 5.20.

in the

figure
was

spring
to

only

in

compression.

stiffness observed in tension

attributed

numerical

causes.

The
the

global cyclic

behaviour
as

as

shown in

figure 5.16 is

characterised

by

the

shape of dynamic

hysteresis loops,
not

defined in

figure 3.18.

It is of importance that the

shape of these

loops does

deviate

too

mueh from the

experimentally

obtained one, since in

NUMERICAL EXAMPLES

113

1.00

0.75

/"~7f
-

0.50
0.25
o CD

mY

Lr^^^^^
/

^T^^ /
/^-^
Y

JZ
in

o in

-0.25 -0.50 -0.75 -1.00

/
/

^^^^

CD
-

-0.04

-0.03

-0.02

-0.01

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]
test, ctd
=

Figure 521

Fourth floor horizontal

displacement versus base shear, cyclic

0.8

2.0
1.5 1.0

r-~

~~~7l
/

\
1

1
I s

[MN]

p cn o p In

/ /

Force

*r

^^^'^^^

-1.0 -1.5
-2.0

tC^^^

I
1
1

-0.01

0.01

0.02

0.03

0.04

Elongation [m]

Figure 5.22 Spring force versus elongation for left vertical spring offirst storey macro
element, cyclic test, ctd
=

0.8

problems the cycle

in

area

within these

loops provides
to
as

a measure

of the

dissipated energy for the

The most

question,

referred

the

hysteretic damping.
the "fatness" and
an

important
the

characteristics of the

hysteretic shape

are

"pinching" of
on

loops.

Among others Saatcioglu [Saat91] provides

extensive discussion

this

topic.

114

CHAPTER FIVE

1.00

0.75
0.50 0.25
o
w

0 -0.25

%
o

co

-0.50 -0.75 -1.00

-0.04 -0.03

-0.02

-0.01

0.01

0.02

0.03

0.04

Fourth floor horizontol

displacement [m]
test, ctd
=

Figure 523 Fourth floor horizontal displacement versus base shear, cyclic

12

2.0 1.5
1.0
-

"

11

s\
/
/

~~

/1
/ l

\
\

[MN]

p cn o p cn

Force

/ //
-

^y^

-1.0
-1.5

I //

mr^^

Z''^
J 1......
1

-2.0

-0.01

0.01

0.02

0.03

0.04

Elongation [m]

Figure 5.24 Spring force

element, cyclic test, ctd
=

versus

elongation for left vertical spring offlrst storey

macro

12

In the model

presented in chapter three of this report both these characteristics of the

are

hysteretic shape

mainly determined by

the factor ac\, described in

are

chapter three
the
chosen

as a

fraction of the tensile

yield force level,

As
oceurs at a

at
a

which flexural cracks

closing on
was

com

pressive
meaning
to

branch

during unloading.

default value, this factor

as

1.0,

that crack closure

compressive force equal to the yield force.

was

Li order

study

the influence of this factor, ac/

varied to 0.8, and 1.2. The results for ac\

NUMERICAL EXAMPLES

115

equal to 0.8
zontal

are

shown in

figures

5.21 and 5.22.

Figure
to

5.21 shows the fourth floor hori

displacement

versus

base shear, and

figure

5.22 shows the left vertical

spring

behaviour.
5.24.
It is
seen

Corresponding results
from

for ac/

equal

1.2

are

shown in

figures

5.23 and

figure 5.21

that

a more

pinched shape

of the

hysteretic loops is obtained

the last

with smaer values for ctd

The

loops

also tend to become somewhat triinner around the

origin.
upper

Less

pinching

is obtained in
case

figure 5.23 (only during

was

cycle,

seen

in the

right corner).

In this

crack closure

barely reached for atf=l.2.

No

particular attempt
would not

is be made to

adjust good

the

input properties

so as to

obtain
a set

an

Optimum agreement

with the

experimental
dever

data shown in this results in other

example.

Such

of

properties

necessarily

comparisons. Rather,

the

general effect of some important properties were shown, with experimental results
the lower stories of a structural wa
It
was seen
as a

from

basis.
to

that the ratio of

as

yield stiffness

uncracked stiffness is

important

for the

nonlinear behaviour,
which however is obtained
tance

weU

as to some

degree the ratio of cracked to uncracked stiffness,

one or two

only important

in the first

eycles. proved

Both ratios must be

to

mainly empiricaUy.
a

The cracked shear behaviour

be of lesser

impor

for

waU of this

configuration.
flexural cracks close

For the

cycc behaviour the points where global hysteretic loops.

largely

determined the

shape

of the

Reasonable agreement could be obtained between both static monotonic and static
set

experimental

data and

analysis

for

cycc behaviour, and this mosy with the chosen cracking

appears
to

default

of

input parameters. Only

the level of shear

be somewhat lower

for the

experiment (around 0.40 to 0.45

MN

compared
model

to

0.57 MN for

analysis),

which

did not influence the

good overaU agreement significantiy.

tests for the macro
are

Thereby, the reabity

directed towards the

completed and the focus wl be

use

Performance of capacity designed budings by

of this model.

5.4.2

Multi-storey wa
of the struc

The

foUowing numerical tests should be seen as dynamic Performance tests

capacity designed
structural wa

tural behaviour of a This action is

buding subjected to

seismic action.

numericaUy apped as a ground motion history.

eight-storey buding presented
to

The structural wa of the

in section 5.3.2 is used here

for the test. This waU

was

designed

resist a the horizontal action of the

buding,

and

116

CHAPTER FIVE

the

sght

moment

resisting

effect of the

gravity

load dominated columns is

neglected

in

the

analysis.

We wl

perform analyses

of the walls for the both

same

designs corresponding to designs,

the difference

restricted and fll

ductity.

The geometry is the

for both

being found in the wa reinforcement.

'

4.0

4.0
i;<"i:i::f

4.0

m
r

H
32.0
m

Elastic region

4.0

m
r

4.0
t.inl(i.,il.i,Oi.fl

Plastic

hinge region

4.0
Uli""
"

2.0 2.0
2.0

m
m m

2.0

m
1

,t

Figure 5.25

Numerical model

of capacity designed eight-storey wall

One mesh

only

wl be used for each

design.

The mesh

corresponds to

the

one

used

by

Bachmann et al in [BWL92b]. With this mesh, the four lowest

eigenfrequencies

for the
were

elastic

(sghy cracked)
=

wa without
=

stiffening beams
=

and nonstructural elements

found to be fi

0.67 Hz, {2

4-0 Hz, f3 found in

9.8 Hz, f4

16.7 Hz.

Eigenfrequencies

for

different states of where it may be

damage

are

Appendix

E. The mesh is shown in

over

figure 5.25,
as

seen

that the

plastic hinge, sti^tching

macro

the

height Lp taken

Lw

6.0 m, is discretised into three

elements.

The
to the

ground motion in form of an artificiaUy generated acceleration history, compatible design spectrum [SIA160]
figure 5.26.
In

SIA

for seismic

zone

3b and for medium stiff ground is

the

shown in

figure 5.26a is shown the time history of

The strong motion

ground

acceleration of 10 seconds

length.

phase
to

lasts about

seven

seconds. In

figure

5.26b is shown the

and for five percent

design spectrum according

the SIA 160 for medium stiff

ground

damping.

NUMERICAL EXAMPLES

117

The

input properties

for the

macro

model,

as

described in

Appendix A,

are

summarized in table 5.3.

Cross sectional area, A

Cross sectional moment of inertia, /

1.8 m2
5.4 m

Young's modulus for smaU strain concrete, Ec

Ratio of cracked to uncracked stiffness for vertical

34000 MPa

spring, Qfcr spring, ay

=

0.7 0.01

Ratio of yielded to uncracked stiffness for vertical Yield moment, Mv,

plastic hinge, (acc.

to

Appendix C), (|Ia

3)

7.9 MNm

Yield moment Afv, plastic hinge, (acc. to Appendix C), (ha

Shear force
at

5)

4.6 MNm
3.0 MN

shear

cracking, Vc,

acc. to

expr.

(3.49)

Ratio of cracked to uncracked shear stiffness, cts

0.16

Table 5.3 Macro model

input properties for eight-storey wall

data to compare the

It is evident that there exists

no

experimental

foUowing

results

with, but instead

the
a

one must

consider the results in view of the intentions which

were set at

design phase.
subsequent

We divide up the treatment of capacity

are

designed buding
buding.
This is
a

so

that at first
in

number of numerical results

shown for the selected

performed

the

and final part of this

chapter. Then,

in

chapter six,
are

number of

more

fundamental
on

questions regarding

the

capacity design

of was

treated, partiaUy based

results of the

examples tested in

this

chapter.

We

direcy

deal with the nonlinear

dynamic behaviour of the capacity designed eighton a

storey wa. The foUowing numerical results base

time

history analysis, performed

input
is

by

direct time

integration,
acceleration

as

mentioned in section 5.2. The

of

given by

the 10
12

second

ground

history

figure

5.26.

Computations were performed for

a

seconds. The time

integration

utilises time increments of 0.01 s, and

residual force

tolerance of 1/100 of the maximum element forces. A maximum of six iterations per increment
were

equibrium
found to be

allowed, whereby
of 2%

two to

three

were

normay

suffieient

Sght Rayleigh damping

(at 0.5 Hz and 5 Hz)

was

prescribed.
be

The first results from the nonnear time horizontal

levels in The
oceurs

history analysis

to

displayed

is the

displacements of the
For

roof level

(ninth floor),
a

shown for both

design ductity

figure 5.27.

comparison

purposes,

linear elastic Solution is also

plotted.

major difference with respeet to the elastic linear Solution hes in the yielding which
for both

capacity designed

walls after around

seven

seconds.

118

CHAPTER FIVE

2.00

-iiiiir

-iiiir-

in

1.00

c o

-1.00

<

-2.00

4 Time

10

[s]

a) Artificially generated ground acceleration history

Spectrum
SIA 160

of

generated ground

motion
~~

Design spectrum

-i

**

1.0
f

10.0

100.0

[Hz]
stiff ground and five percent

b) Design spectrum according

to

SIA 160 for medium

damping
Figure 526 Artificially generated ground motion
It may be in the

seen

figure

that the difference between both

ductity

levels is

not
a

particularly large. This

hinge.
The

may be

explained by the fact that die fll ductity design gives

damage oceurs
in the

softer system, for which the first mode is excited less after

plastic

global displacement

ductiUties

set

at the

design phase cannot be verified direct-

NUMERICAL EXAMPLES

119

0.15

-j

ri

"

-1

Linear Elastic Refe

rence

Solution
_

rf
g
E
u

0.10

Design Design

p.

= =

3
5
A\

p_

0.05

j
A
r

j\\
i

A
i \/
'

s
Q.
w

Vv if^ \\ J
\ V

i/\v/ \
Xj

ji
f
/

/P
h

\\

\\JrJ
\
v'

-0.05

i 1i

ArA

\y/

V.

Vi
-0.10 -0.15 0
1 1 1

...

...

6
Time

10

12

[s]
roof displacements

Figure 5.27

Horizontal

0.15 0.10
c

>'r

"

--

T
A
=

0.005

...

0.01 0.02

s s

AT

0.05

J\,\ A
V

/ \i
ffl

ftt
-

Q.
Vi

-0.05

r\\

-0.10h
-0.15
0
i
i i

'

6 Time

10

12

[s]
horizontal

Figure 528 Influence oftime

fll ductility

increment

length,

roof displacement, design for

ly

from

figure 5.27,
be close

since die roof displacement

at

the

onset

of

yielding at

the

base,

was

found

to

to zero, or even

negative due

to

the influence of higher modes, which

would give infinite

displacement ductity values. length, the roof displacement of

s, 0.01 s,

In order to estimate the influence of the time increment the fll

ductility design is

shown in

figure 5.28 for increment lengths of 0.005

120

CHAPTER FIVE

E
z

c o

E
o

Time

[s]

Figure 5.29 Overturning moment at wall base versus time, restricted ductility design

E
z

E
o

Figure 530 Overturning moment at wall base versus time, fll ductility design
and for 0.02 It is not

s.

possible

to

distinguish

between the three solutions until after

about

eight seconds, where sght differences occur. No essential difference in maximum

The solutions for die
two

displacement can be observed.

smaller increments appear closer

together indicating convergence.

The nonlinear behaviour at the base of the wall may be

displayed through
moment at

the

over

turning moment and

shear quantities. We start with the

overturning

the base,

NUMERICAL EXAMPLES

121

12.0

-i

r..,,...,

Right
Left
c

curvature

urvature
-

10.0

E
Q

8.0 /

z*

b.O

Q
<D
l_

4.0

O
>
i_

2.0

r'
l l l

/,
iiiji1_
ii

...

10

12

Time

[s]

a) Restricted ductility design

12.0

i|iiir

Rinht

mirvntii

ro

11
o

Left curvature
10.0

E
a>

8.0

f'

6.0

Q
ffl
u

4.0

o
>

.1
l

2-

10

12

Time

[s]

b) Fll ductility design

Figure 5.31 Curvature ductility

shown

versus

time

versus

time in

figure 5.29 for the restricted ductility design and

The flexural demand at the wa base M., taken

in
as

figure 5.30 for

the fll

ductity design.

ykME (where
equivalent
we

yR is the resistance factor of 1.2 and

ME

is the moment

at

die base from the

static force calculation) is shown in the

figures.

With the roof

displacement known,

122

CHAPTER FIVE

12.0

i'ir~

10.0

ay ay

0.01

0.02

E
<o

8.0

6.0

o
o
l_

4.0

>

2.0

6
Time

10

12

[s]

a) Restricted ductility design

12.0
-o c o

,i.i,i
et
_

.1,111,

1"

"

"1

""

T"

'

0 01 0.02
-

10.0

ay

E
<B

8.0

(-t

6.0
Q
<D
1_

4.0

"5
>
i_

z.

2.0

llllll

i.

..

i...

6 Time

10

12

[s]

b) Fll ductility design

Figure 532 Influence of yield stiffness on

may here estimate the P-A effect

curvature

ductility
2 MN
one

to

be about 0.1 meters

(displacement) times
as

(gravity

load for upper stories)

0.2 MNm, i.e.

conservatively estimated

around

percent of the base moment which may be neglected.

NUMERICAL EXAMPLES

123

The curvature ductilities

as a

function of the time each

are

given in figure 5.31

show the left and

and 5.32 for and left

the lowermost element of the

curvature.

plastic hinge,
and 5.32

plot separately showing right

The

plots of figures 5.31

time.

separately

right curvature

demand

versus

The any

ductity

is defined

as

based

on

the maximum curvature which has been reached at

given
to

occasion. These values

outer

were

obtained

using the default yield stiffness factor cty

equal

0.01 for the

on

vertical

Springs.
0.02

It is clear that the

yield stiffness has

seen

some

influence

the maximum curvature

ductity reached. This

are

influence is

in

figure
clear

5.32, where the results of cty taken

as

shown. Maximum values from left and

has
a

right

are

compared. It is

seen

from

figure 5.32 that higher yield stiffness

reducing effect
In order to get

better view of how the

bending

moment

and curvature relaxe, and

distribute
for each

over

the lower part of the wall, three

moment vs. curvature

plots

are

displayed
plot on
ele

ductility design. Figure 5.33

relation in three

shows for the restricted describe this relation

ductility design the moment

at

curvature

plots which

(starting

with the

the bottom of the

ment

page):

the lowermost element of the element from

plastic hinge, the uppermost

of the

plastic hinge (third

the

below), and the lowermost element of the

elastic
In

region (fourth element from below).

figure 5.34,
from

corresponding quantities
the

are

shown for the fu

ductity design.
degree,
some

It

is

seen

figures 5.33 and 5.34 that the curvature clearly decreases towards
bending
State at
moment

the top of since the

the

plastic hinge, however,

in
a

decreases

to a

lesser
seen

plastic hinge is still

takes
The

yielded

its upper end. It is also

that

yielding

place

at

the lower end of the elastic

are

region, especially for the

fll

ductity design.

impcations of this

treated

more

fuUy in chapter six. global displacement ductity proposed

in

Since
curvature

relationship

between chosen

and

expected
a

ductility

in the

plastic hinge

is

[PBM90] based upon

static

analytical example
noted that the

it would be

possible to

compare this

proposal to the curvature ductity

obtained from the time

history analysis

in order to check the agreement. It should be

plots only displayed

the curvature
over

ductility for the lowermost element of the plastic hinge height. Due
average value
to

plastic hinge, while [PBM90] defines it

discretization of the exhibits
a

die entire

the

plastic hinge into

than

three elements, the element dosest to the base

a

larger curvature ductility

distributed curvature

value

suggested

as an

assuming
chapter

uniformly
six.

ductility.

This

matter

is also discussed further in

S'
Moment
Moment
I
M
-

I
l
MM-- Ol O Ol O
M

Moment

[MNm]

[MNm]
[MNm]
I I
Ol
o

Ol

tili, - I MM l I
Ol cn Ol --MM Ol O O Ol

--MM

Ol

Ol

Ul

oi

p
p
O

oi

p
Ol

in

--MM

Ol

Ol

lb b
b
I

b b
b b
lob
Ol

b
b

5 ?
o

I o

>M

3 3

3 3

^^
1

Moment
Moment

[MNm]
[MNm]
Moment

l l

[MNm]
Ol --MM o Ol o Ol o

s 2
o o

I
MM-- Ol O Ol
~

I
l

I
I
--MM O Ol O Ol MM-- Ol O Ol O
,
,

M Ol

Ol

I I
l

Ol

I I
Ol

--MM O Ol O Ol

1 o

o
o

lob b
o

in
T
T

1-

1 1
i

lob' p 1 I cn

5
tn
C/i

1 o

1
-

I
o

o gm

3 3
o

ET
c
ro

o
r^

3
tt
!t*
-

^
3 3
<*
Jt *o
M

3
^
o

i_Up
a
JH Ci'
a-

.
>t <*

oo
o o

Sl *9
-

\
(*
N

o a

1 1
en

1
o

O
Ol

Oq 2.

CVI

126

CHAPTER FNE

In order to foow die nonlinear

spring

behaviour during the time

history analysis, spring

the

spring

force

versus

the

spring elongation

is shown for the left vertical

as

of the

lower- and uppermost elements of the

plastic hinge

we

as

for the lowermost element

of the elastic

region,

in

analogy with figures 5.33

spring
In these two
some

and 5.34.

Figure

5.35 shows the

behaviour for the restricted

ductity design,
more

and

figure

5.36 for the fu

ductility design.

figures

it is

clearly

seen

than in

figures 5.33
elastic.

and 5.34 that there is indeed

yielding in the region

the

intended to remain

For both the restricted and die fu

ductity design,

point where cracks close on the acting

the

unloading branch,

was

taken

as

four times the

yield force in tension, i.e. ad=4, giving a

on

closure force level somewhat below and above the effective normal force

wa, for the restricted and fu ductity level respectively.

From

figures 5.33, 5.34,

as

well from 5.35 and 5.36, it may be

for the fu

seen

that the

unloading wl be foow a flatter curve

ductity design
far the

than for the restricted

ductity design.
PhysicaUy,
in this may be

thought of as reasonable in
more

so

as

compressive yielding
due
to a

unloading should be
weU due to the
to a

somewhat

pronouneed

for the fu

ductity design

smaUer vertical reinforcement

as

area over

which the compressive forces may be transferred that has occurred in tension before of

as

larger plastification
more

unloading.

This leads
curvature most

somewhat

"pinched" shape
be 5.34.

hysteretic loops of the

seen

moment vs.

for the fuU

ductity design, as may

clearly

e.g. in the

plot for the lower

element in the

plastic hinge of figure

The shear force at the wa base is restricted

displayed

versus

time in

figure 5.37
In the

for the
the

ductity design
at

and in

figure

5.38 for the fu calculated

as

ductity design.
=

plots,

shear force demand

the wa base is die

Vw,

Vw

oojb.^ (where <J>oa

and

is the
at

overstrength factor, t,

dynamic magnification factor,

force calculation,
as see
see

VE

is the shear force

the wa base from the

equivalent

section 5.3.2) is shown

sum

as a

dotted ne. The shear

capacity VR, determined

the

of concrete and steel

contribution based on the

the

capacity design method,

section 5.3.2, is also indicated in

plots.
seen

It is

that the

capacity is mueh larger than the equivalent static force. This is mainly
length (6.0
m

due to the

large give

wa

horizontaUy)

in combination with the minimum

requirement
contribution
this

for horizontal reinforcement of 0.20 % which

together
a

with the

concrete

around 2.7 MN

capacity,

see

section 5.3.2. For

wa with less

length
It

capacity
seen

would be smaUer,

although

the minimum

requirements would be fulfled.

an

may be

that for both the restricted and fll

ductity designs,

effective shear force

NUMERICAL EXAMPLES

127

2.0

1l

I
O

-2.0

S.

-4.0

k /

'/

-6.0

-8.0

-0.002

0.002

0.004

0.006

Elongotion [m]

Lowermost element of elastic

2.0

region
1

-2.0

ff

m-

-4.0

//

-6.0

-8.0

-0.002

0.002

0.004

0.006

Elongotion [m]

Uppermost element ofplastic hinge

-0.002

0.002

0.004

0.006

Elongotion [m] Lowermost element ofplastic

hinge

Figure 535 Left vertical spring behaviour, restricted ductility design

128

CHAPTER FIVE

2.0

-2.0

i
-

-4.0

-6.0

-8.0

'

'

-0.002

0.002

0.004

0.006

Elongation [m]

Lowermost element of elastic

2.0

region
1

k
-

//IJ/I

-2.0

-4.0

-6.0

-8.0 -0.002
0

0.002

0.004

0.006

Elongotion [m]

Uppermost element ofplastic hinge

2.0
l
l

-2.0

l^^ ^s i

'//

"^

m^l l

-4.0

1
1

m-^^

-6.0

-8.0 -0.002

.....I

..

0.002

0.004

0.006

Elongotion [m]

Lowermost element

ofplastic hinge

Figure 536 Left vertical spring behaviour, fll ductility design

NUMERICAL EXAMPLES

129

Figure 537

Base shear, restricted ductility

design

Figure 538
is reached which for fu
close to the

Base shear, fll

ductility design

come

capacity (2.3

MN for restricted

ductity

and 2.1 MN assumed

ductity). Due to conservative estimates on the concrete contribution it is

fall in shear.

that the wall would suffer extensive shear would

not

cracking during the simulated earthquake, but it

based

Since the time

history results
to

so

far

are

on an

artificially generated ground

same

motion, it would be of interest

exarnine the Performance of the

buding subjected

130

CHAPTER FIVE

4.00

Vi

2.00

c o

o o

2.00

<

-4.00
10 Time

15

[s]
recorded at Tolmezzo, Friuli, 1976

Figure 539
to a recorded
a

N-S component

of ground motion

ground motion. For this purpose we repeat some of the computations, using
at

N-S component of the ground motion recorded

The
a

Tolmezzo, during the 1976 Friuli

5.39. The record is of 20 seconds
at

earthquake. length,
with

ground

motion is shown in
seven

figure

strong motion phase of

seconds, starting

around 3 seconds.

It should be noted that the maximum

factor 2.3
same

ground

acceleration is around 3.6 the SIA

m/s2,

i.e.

higher than
as

the

ground

acceleration
are

compatible to
to

design spectrum.

The

walls

used

previously
was

subjected
realistic

the recorded

ground motion, without

check for the

being redesigned. This

two

done

intentionally

so as to

get

Performance

designs,

when

subjected

to a

ground

motion with

significantly higher

acceleration levels.

As

first

plot,

the roof displacement versus time is shown in

m

figure 5.40.
an

A maximum

displacement

of around 0.13
=

for the fll

ductility design gives

to

average inter storey

drift of 0.13/32.0

0.004, i.e. 0.4 percent, compared

around 0.3 percent for the SIA

compatible ground
Rather than

motion

input,

both

being relatively moderate values.

limit the results to

repeating

all the

previous plots,

we

some

critical
versus

quantities,

of which the base shear increased most

dramatically. The

base shear

time is shown in

figures

5.41 and 5.42 for both

designs respectively.

NUMERICAL EXAMPLES

131

0.15 0.10

'

'

11
c
o

1*1

r.

0.05 h

l\

v.

1 1 /

E
tu

1 \ \i 1 11
\

/Ma1.

/\V

0
\

a.

1
*

U
\

IV
1
'

l'
M
y

Vi
3 5
i

\\ /

-0.05
o o Cc.

VT

\f

Linear Elastic Reference Solution

-0.10

--

Design Design
i

pk
fx.
. i

-0.15 0

6 Time

10

12

[s]
roof displacements

Figure 5.40
It is which

Horizontal

seen

that for both

designs,

an

effective base shear of around 2.7 MN is reached, and

means

that the shear

capacity is fully utised,

reaUy is,
that
a

depending on

how conservative

the shear capacity estimation

shear faure may not be excluded.

It should be mentioned here

not

again

that the shear behaviour of die

macro

model does

simulate shear failure, since this is

a

partly
the

not

desirable and

partly

difficult to obtain

realistically in
shear

simple numerical model. Rather, by

user.

cracked shear stiffness is used, and the Realistic

capacity
here

should be estimated

modelling
Some

is obtained

by

the

model until the

region

where the shear

capacity is reached.

thought

should rather be

given

to

the

question

about how reastic is the estimated shear capacity.

For the estimation of the concrete contribution to the 2.7 MN

capacity,
was

fu normal

force of 4.15 MN

(including

dead and live load

tributary
1.2,
a

to

the wall)

used, without

multiplying

it

by

the resistance factor yR in the literature (or

equal

to

as

is sometimes done in reduetion factor

capacity
taken
as

design examples
If, however,

dividing by

strength

(j>

0.9) and which may be rather unconservative.

we

further reduee the effective normal force

a concrete

acting

on

the wa to 90

percent of its value, this results in

shear

strength of 0.86
when

MPa rather than 0.91 that

a

MPa, and
failure is
like the

total shear

capacity

of 2.56 MN rather than 2.7 MN,

indicating
a

shear

quite likely for both

used here and
a

the

ductility designs,

subjected to

ground motion

one

normal force

slightly smaller than that assumed.

132

CHAPTER FIVE

Figure 5.41

Base shear, restricted ductility

design

Figure 5.42
The curvature

Base shear, fll ductility

design
also

ductility demand obtained for die recorded input was input.

The curvamre
versus

clearly larger
the

than for the SIA

ductility demand for the lowermost element in

plastic hinge

is shown

time for both

designs

in

figures

5.43 and 5.44,

respectively. For the restricted ductity design, initial yielding in right curvature occurred
during reloading
ductity.
More from

yielding

in left curvamre which resulted in

large

curvamre

pinched hysteretic

behaviour for the fu

ductity design delayedthe

NUMERICAL EXAMPLES

133

12.0

>

Right

curvature

Demand Ductil ty
|
D

Left curvature ob

00 b

/
r-1
-

4.0

2.0

t"

i
0
.. i

10

12

Time

[s]
restricted ductility

Figure 5.43

Curvature

ductility,

design

12.0
-a c o

iii,..,

,111,1.1,1.' i

Right
-

curvature

Left curvature 10.0

E
o

8.0

u 3

6.0

I
.
.

o
a>

4.0

./-

*""

"5
>

i
2.0
i
-

,.i

...

...

10

12

Time

[s]

Figure 5.44

Curvature

ductility, fll ductility design

yielding during reloading

The results

and thus mited the resulting curvature

ductity.
to

displayed

from the

eight-storey building
the

were

intended

give

basic
serve

over-view of the
as a

analysis capacity of the numerical model for such problems,

for the discussion
of the
macro

and to

background
the
tests

on
are

capacity design
in

method in

chapter

six.

Thereby,
to some

model

concluded and the rest of this

chapter is devoted

basic

tests

of the micro model

developed

chapter four.

134

CHAPTER FIVE

5.5 Micro Model Results

In this section the micro model which

was

developed
will
use

in

chapter four is subjected

same test

some

reabity
macro

tests. For

the sake of

simplieity

we

the

specimen

as

for the

model, i.e. the three-storey specimen tested by Vallenas

et

al in 1979. since the estab-

We wl

mainly

limit the numerical tests shown here to

as

reliabity tests,

shment of meshes

we

as

die numerical effort

are

both

quite considerable
wl

when this A

model is used to simulate tau

multi-storey

structural was

subjected to ground motion. complement

few results from numerical simulations of the

eight-storey building

the

reabixy tests.

5.5.1

Comparison

with

experimental

results

The

specimen
set

and force

application
was

are

the

same as

described for the

macro

model. A

basic

of model

properties

established from which deviations

macro

are not

made unless

specificaUy stated, in
table 5.4.

accordance with the

model. The

properties are summarised in

Young's modulus for concrete

Poison's ratio for
concrete

27900 MPa 0.20 0.0001 211000 MPa

444 MPa

Cracking strain Young's modulus for steel

Yield strength for boundary element reinforcement
Yield strength for web reinforcement

507 MPa

Strain at seized tension

stiffening
terms

0.002
of elastic modulus

Strain hardening ratio for steel in

0.01

Table 5.4 Input properties for Micro model

test

examples

The hear retention upon

cracking is modeed according to [Rots84] finay,

mesh C has 170
was

Three meshes

are

tested, shown in figure 5.45. Mesh A consists of 28 elements

regarded

as a

minimum mesh, mesh B has 60 elements, and

were

elements. Forces
at

discretised and

applied at as follows.

The shear force

apped

a nodes of the uppermost elements. The vertical forces

were

(gravity

and vertical force

couple)

applied

at

the four nodes of the uppermost

corner

elements.

NUMERICAL EXAMPLES

135

V /*Y
-

^C

:;';

YP

Vi
:>,
'

s
'

,'

<-

>;:,

v-

'.

'

':

*':
<
,,

'>
^\

\-

^/

^
'%

:"

,"
'
.

:"

J-**:

1.
"

o-s >.r
A

^^^

'rYj:
v

v;* 'k*j
"

%? 'o!';

....

V
^

A
s.
.'.

ii
i '/r

-.;

<f'
"

A ?'

-,

~!
5

m
*_>

r-

-.

<
,

x*'m<3.

$
5.
-..

#
Y-'i
rs,/,M
'
*

"

-J^

*?
.V

..

/ \'A
/r

*: ?"
y-

i
Vf

^ i'i
*
,4-

^
/i
'

>'

Y''

4
?
v>

'<

'

/ '

$
\
v

fr *!

'"<

^'

-S

'.$.

;..
^

'

0^;
a) Mesh A

Ayy-

*%

tr-'

..

:&
:,-.

v<& -%
$

:#'
*

''

'*%

i>

Y.

'

%' i"
r-

* 9

i?

*f
--

H
-

b)MeshB
Figure 5.45 Meshesfor micro model

c) Mesh C

1.0 0.9

.-1

'

0.8

--

^0.7
2
1-1

/-

#y'*

0.6
-

Mesh A Mesh B
Mesh C

ja
to

0.5
0.4

0 10

0.3
-

0.2

0.1
0
j
1

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

base shear Figure 5.46 Fourth floor horizontal displacement versus

The fourth floor horizontal

all three meshes. It is
seen

displacement versus base

subsequent plots mesh

shear is shown in

figure 5.46 for

are

from

figure 5.46 that the results of meshes

B and C
as a

closer,

indicating convergence.

For the

B wi be used,

compromise

between aecuracy and numerical effort.

136

CHAPTER FIVE

ft

2.5 MPa 2.8 MPa

ft
ft

3.5 MPa

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure 5.47 Influence of concrete tensile strength

The influence of the concrete tensile

strength

is shown in

figure 5.47, using mesh

seen

B.

This

strength

is varied from 2.5 MPa to 3.5 MPa

(default value: 2.8 MPa). It is

where the

that

the influence is

mainly visible in

the

pre-yielding region,

higher tense strength opposed

gives
the

somewhat

higher global strength.

It should be noted that the micro model

essentially simulates uncracked tensile behaviour, via its

macro

skeleton curve,

as

to

model, where cracked flexural behaviour

was

was

assumed

at

the start of the

analysis,

whereas the shear behaviour

assumed elastic.

The next parameter to be

analysed is

the influence ofthe strain

hardening ratio for the

to

reinforcement, shown in figure 5.48. The ratio of the yield modulus

modulus is here varied from 0.005,
over

the elastic steel

0.01

to

0.02. The

global

effect is

essentially

proportional to
ratio represents

the set parameter,

more

as

may be

seen

from the

figure.

In the micro model this

direcdy than

in the

macro

model the

slope of the yielded branch for

the reinforcement

only.

Furthermore, the influence of the tension stiffening effect is shown in figure 5.49. The
strain
at

which the tension

stiffening stops acting, i.e.

over

where

zero

tensile
at

stress

is reached
to

for the concrete, is varied from 0.001

0.002

(approximately

steel

yielding)

0.003. For the low value of 0.001, the

increase may be observed. It is due

typical temporary yielding foowed by a stiffness

the temporary
over

to

negative modulus in the tension

stiffening phase, appearing simultaneously

post-tension stiffening phase
with
zero

several

integration points,
to

foowed

by

modulus, which is stiffer than the negative

modulus, compare figure 4.14. The value of 0.002 appears

give

reasonable result

NUMERICAL EXAMPLES

137

ay
aY

0.010 0.020

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]
ratio

Figure 5.48 Influence of strain hardening

1.0
0.9

0.8
rr

--

mr

"z
-

0.7 0.6 0.5

0.4

'/^
~

'
--

e0

0.001

O
CO

,/
r y

'

fr'

e0- 0.002 e0
=

//

aj (0

0.003

0.3

0.2
0.1

0 L

_.

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure 5.49 Influence of tension stiffening

The next

quantity to be analysed is

the shear retention. The default shear retention used

based upon

for this micro model is the

expression by [Rots84]

experimental expression
see

work

by

Paulay

and Loeber,

see

equation (4.16).

We will here compare this

with die

expression by Bazant and Gambarova [BG84] which is of a simar nature,

(4.17), and with the expression by Pruijssers [Prui85], equation (4.18).

equation

138

CHAPTER FIVE

acc.

to to
to

[BG84] [Rots84] [Prui85]

acc.

acc.

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure 550 Influence of shear retention G

The influence
two

on

the

global

behaviour is shown in close ([Rots84] and

figure 5.50.

It is

seen

that the first

expressions
a

are

relatively

[BG84]), whereas the expression of

to

[Prui85] gives

stiffer

global

behaviour. This is due

the fact that in this

a

expression
a

small ratios between normal strain and shear strain

give

larger

retention factor for

given normal strain. Since

of 19

there is mueh

bending (i.e.
a

normal tensile

stress)

in the

computed example this effect shows up clearly. Further,

mm was

maximum aggregate diameter

for the 1:3 scale

assumed, which may be slighy

to

too

large

specimen.
curve
one

Reducing this

12

mm

gave

somewhat lower stiffness, still above the

for

[BG84] however.

It may be concluded that the

to

simplest

of the
to

expressions,

the

by

[Rots84] appears

give

reasonable results

compared

experimental

data in this

example.
Some attempts with
more a constant

shear retention factor showed that factors of 0.5 and

gave

mueh

too

stiff Solution, and lower factors gave considerable

to

equibrium
the tensile
shown in

iteration

problems
be

due the

the sudden

drop

in shear stress when

passing
the

strength. Therefore,
the

gradually decreasing nonlinear functions,

the influence of the

as

ones

plot,

seem to

preferable. Young's modulus of the reinforcement steel,

the

We will

now mm to

discussed

at

the end of

chapter four. Modifying

must

Young's modulus allows

better

Simulation of the additional stresses which

across

be carried

by the reinforcement bars

arise from

cracks. As discussed in
stress

chapter four the additional stresses in the bars

concrete

tense

carried in the uncracked

sections.

NUMERICAL EXAMPLES

139

It

was

shown in

expressions

4.21 to 4.23 how

possible
tests,

modification of the steel the

modulus may be

performed.

In the

previous reabity

using

example

with the

three-storey test specimen, no modification of the

In order to

steel modulus

was
a

made.
we

study the influence

in

on

the

global

behaviour of such
to

modification

here

perform

modification of the steel moduli relevant

the

three-storey specimen,

according to the Suggestion

Table 5.4.

chapter four.

The necessary material data is

given below in

Concrete tensile

strength// (tension stiffening stress)

2.8 MPa

Concrete stress between cracks

0.4/,

1.2 MPa

Unmodified Young's modulus Es, for a reinforcing steel

Yield

211000 MPa

strength: boundary elements/web

444MPa/507MPa

Yield strain: boundary elements/web

Vertical reinforcement ratio:

0.0021/0.0024
3.50 %/0.73%
0.73 %/0.73%

boundary elements/web boundary elements/web

Horizontal reinforcement ratio:

Table 5.4 Material data for check of tension

stiffening
not reach a stress

Since the horizontal steel,

in the

as

wl be
we

seen

later in this here

a

chapter, does

vicinity of the yield strength,

as

perform

simpfied modification

of steel

modu,
based
on

follows: In the

boundary

element horizontal and vertical steel is modified

the vertical reinforcement ratio

(overestimates die modification for the horizontal

on

steel). In the web both vertical and horizontal steel is correcdy modified based
vertical reinforcement ratio of the web, which is the
same as

the

the horizontal ratio.

Modification ofthe steel modulus

steel modu,
as

according to expressions 4.20-23 gives die modified

follows:

Boundary element:

fts=^s- ^=

0.035

32 MPa

f.

AAA

32

m + u

=-------=

=226700 MPa

y
Web:

0.0021

/=-2s-=
J"

-^0.074

151 MPa

=-4lA=444 + 151
y
0.0024

274200 MPa

140

CHAPTER FIVE

Modified steel modulus

Unmodified steel modulus

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]
to

Figure 5.51 Influence ofmodified steel modulus due

Thus

tension

stiffening

we

obtained

an

increase for the

to

Young's

modulus for the steel in the order of

226700/211000 equal
274200/211000 equal
to

1.074, i.e. 7.4 percent for the boundary elements, and

1.300, i.e. 30 percent for the web.

The effect

global
a

influence of this modification is shown in

stiffness in the

figure
and

5.51. It is
a

seen

that the

gives

slight increase in

cracking phase

few percent additional

strength. Inspection of figure 5.8 suggests

achieves It
to
can
a

that the influence ofthe modified steel modu

result which is
not

little closer

to

the

experimental data.

however,

be stated that this influence is of signifieant

importance
studied

in relation

the influence of other parameters of the micro model which

were

previously in

this

chapter.

Some of the internal behaviour of the micro model is

presented

here

to

conclude the

monotonic

loading study. edge of the base level is shown in

of the

The behaviour of the concrete and the steel at die left

figures

5.52

applying

to

mesh B. The lower outer

used to obtain the

integration point

edge element
vs.

dosest to the base

was

plots.

In 5.52a the concrete stress

vs.

strain,
across

normal to the cracked cracks is

plane is displayed.

In 5.52b the shear stress

versus

shear strain

displayed.

In 5.52c and 5.52d the steel stress

strain is

displayed for the

NUMERICAL EXAMPLES

141

4.0
r\

3.0
10 to

mr^

H
mt~

2.0

CO
L.

S
sz

i.o

C/l

11I.-.J.11J11

0.000 0.005 0.010 0.015 0.020

0.000 0.002 0.004 0.006 0.008

Shear strain

Normal strain

a) Concrete stress vs. strain

in direction to crack plane

b) Concrete shear stress

across

vs.

strain

cracks

o_

w
co

tn

0.000 0.005 0.010 0.015 0.020

0.0002 Strain

0.0004

Stroin

c) Vertical steel stress

vs.

strain

d) Horizontal steel stress

vs.

strain

Figure 552 Internal stress quantities of micro model, left edge at base
vertical and horizontal bars,
A

respectively. cracking is
shown in

global pattern of the

concrete

figure 5.53. for three

different

shear force levels up until 1.0 MN. In 5.53a the crack pattern from the prototype after

completion of test is
is shown

shown. In the

plots

5.53b-d the crack pattern from the micro model

which
are

using

element CPS4. The

numerically obtained crack pattems,

shown

142

CHAPTER FIVE

a) Test specimen crack pattern

b) Numeric crack pattern

Shear force 0.4 MN

after carried out test (1.0 MN)

HWlwlw)
-^-\

w
^
_

hKxkywK
\

-^\

w
>>

_-

->.\ w

S> V.N \

-~.

w \

-SWiW

\1\

c) Numeric crack pattern

Shearforce 0.6 MN

d) Numeric crack pattern

Shear force 1.0 MN

Figure 553

Crack pattern for mesh B

for mesh B, show reasonable agreement with the the final

experimentally obtained crack pattem in

phase.
the

Thereby,
be directed

reliability
to

tests are

concluded for the micro model, and die attention wl which

direcy

the

dynamic behaviour of this model,

incorporates

the

cycc

behaviour.

NUMERICAL EXAMPLES

143

5.5.2

Multi-storey

wa

A limited number of the results from the

eight-storey
is very

wa shown for the

macro

model,

wi here be
a

repeated

for the micro model. Since the nonlinear time

history analysis using

given
here. In

micro model, well

as

simulating multi-storey was,

only
some
a

comprehensive regarding modelling

of results is
a

as

numerical effort,

mited

comparison

separate subsequent section

and micro models.

results wl be

direcy compared
shown in

between the

macro

The micro model mesh for the

eight-storey waU is
distributed
to

figure 5.54, consisting of

masses were

68 elements.

Gravity

loads

were

floor nodes, and floor

lumped

into two

edge nodes at every floor.

Free Vibration data is

given in Appendix

E.

4.0

Vertical reinforcement for

plastic hinge elements:

Restricted

4.0 m
f'%

displacement ductility:

Py
4.0 40 4.0
m

Py=
0.26'

Py=
0.26%

Py

1.2%

12%

J05J.
Fll

1.5

X.

1.5

M
Py

4.0 m

displacement ductility:

2!m
2.0 2.0 2.0
m m m

Py=

Py=

Py=
0.21 %

0.46 % 021 %

0.46'

JOS)

1.5

J.

1.5

Jpj].

J,6-0mJ,
Figure 5.54
The Micro model mesh for

eight-storey wall

edge

elements of 0.5

lengi correspond
elements

to the confined end zones of the was

containing
over

flexural reinforcement For the restricted

ductity design
were

the

edge

elements

the

plastic hinge (four

vertically) only

given

1.2 percent vertical

reinforcement, and for the fu ductity design 0.5 percent. The artificiay generated

ground motion compatible

of 0.01 seconds

to

the SIA spectrum

wl be used here. A time increment

only

wl be used.

Rayleigh damping of 2% (at 0.5

versus

and 5

Hz)

was

used.

The horizontal roof

displacement
macro

time is

displayed

in

figure

5.55 for both

designs.
ment

The

finding

from the

model results, that the difference in roof

displace
the micro

between the two

designs is

not as

large

as

expected,

is here confirmed

by

model.

144

CHAPTER FIVE

0.15

->r

~i''r~

Linear Elastic Reference

Solutlo/i
'^

0.10
c
<D

Design Design

\ik
\x.

.'""

5.,

0.05

Q.
w

-0.05
o

oc

-0.10
-0.15 0
j L

6 Time

10

12

[s]

Figure 555 Horizontal roof displacement versus time Many quantities

such

as cross

sectional forces and

to

curvature

ductility

are not

defined

for the micro model in

out

a manner

comparable

the

macro

model and wi therefore be left

as a

of the discussion. The roof displacements wl suffice

smaU demonstration of the

nonlinear

dynamic capabilities
a

of the micro model, and the last section of this

two

chapter is

devoted to

brief comparison between the

models.

5.6

Comparison

between Macro Model and Micro Model

In this section,
macro

some

comparisons

will be

presented

between results obtained

by

the

and micro models.

Although it would be possible to perform a relatively extensive

quantities
for both models, this

comparison
limited
to a

between various result

presentation
at

wi be

few

plots displaying the displacement ofthe eight-storey wa

roof level.

For restricted there is

a

ductility design,

the

comparison is
macro

shown in

figure

5.56. It is

seen

that

reasonable agreement between

and micro models.

a

During

the last few

seconds of the time

history analysis

the

macro

model exhibits

sghy longer period and

in

somewhat smaller
The
same

amplitudes.
results for the fll
for the restricted

corresponding
as

ductility design
ductility design

are seen

figure 5.57.

The

observations

hold true. For the last three

seconds, however, the micro model clearly shows larger amptudes, indicating more pro-

NUMERICAL EXAMPLES

145

0.15

TT"-TTpT-ri-i-|jr-

-l

r-

Macro Model

If

0.10

Micro Model

-0.15
0

'

'

1-

6
Time

10

12

[s]
ductility design

Figure 556

Horizontal roofdisplacement, restricted

0.15

'

->ir-

Macro Model

"e

0.10

Micro Model

1
E
o

0.05

I
O(0

-0.05

-0.10
-0.15 0
*

'

,*

6
Time

10

12

[s]

Figure 557
nounced

Horizontal

roof displacement, fll ductility design

yielding

than the

macro

model. The somewhat thinner

shape

of die

hysteresis

loops

of the micro model may be

The
reason

responsible
originales

for the evident less

to

damping

in the post
as

yielding phase.
shown in

for this

the concrete

hysteretic

behaviour

figure 4.9 combined with the reinforcement hysteretic behaviour of figure 4.14.
the reinforcement is

Of diese which

two

responsible for the major part of the hysteretic damping,

A further

altogether is somewhat insufficient

explanation is

that the

macro

model

146

CHAPTER FIVE

keeps

the

yield stiffness

set at

the

beginning, also for deep excursions into the nonlinear

may be set.

rnge, which mits the

amplitudes somewhat more than is the case for the micro model
also tested for the

for which

no

comparable specific yield stiffness

die results

A finer overall mesh for the micro model

was

eight-storey wa, but

plastic

did

not

change

significany. display

The

same

holds for

finer discretised

hinge region only.

The fact that both models
a

reasonable agreement in these

highly nonlinear
may be valid for

problems, except for

seen as a

at

the end of the time of the

history

for the fll

ductility design,

further

proof

reabity

of each model. This is

especially

problems which do not have experimental data to check against

Having discussed mainly the numerical aspects of the

conclude with
a

two

models, this chapter will

small overa

comparison

of the two models,

whereby

their

respective

advantages and drawbacks wi be underscored, and die ideal

identified.

use

of each model wl be

The

macro
as

model
in
a

was

from the

beginning of

this

study designed
nonlinear

to use the same

parameters
etc.

design

process, that is sectional forces, curvamre in the used in die

plastic hinge,

Thus it may be

readily

modeUing and

dynamic analysis of Although of

global buding
a

structures

for direct comparison of the mentioned parameters.

relatively simple mechanical nature, it is capable of simulating the global behaviour of subjected
as

structural walls

to

earthquake

action. The

typical structural

wall behaviour

characteristics such

large tensile strains, pronouneed shifting capable

of

of the neutral axis, and

shear deformation may a be simulated

to

relatively we by this model, especiaUy compared

simulating
the

other

macro

models. It is furthermore

typical hysteretic

behaviour under load reversals relatively we. In the present

was within
one

implementation,
building

as

described in the
as

Appendix of this report, several they

mn

may be modelled

long
not

as

in the

same

direction

(essentially two dimensional analysis). Although

this report,
a

described within the framework of

macro

minor test

was

performed

on a

modified

model

prepared

for three

dimensional

analysis (still with planar waU action) planned,

with

satisfactory results.
such
as

A fll three

dimensional version is

and three dimensional

cores

lift shafts should be

possible to model ready.

In the present version the
macro

model is

capable

of

providing

aU desired

cross

sectional

quantities, such

as

section forces and curvature

ductity.

AU these
cross

quantities
sectional
a

may be extracted for any element

along the height of

the wa. These

quantities

are

very valuable in the confirmation of die

dynamic Performance

in

design

NUMERICAL EXAMPLES

147

process and for research purposes. These

quantities,
macro

which

are

obtained

readily,
the

constitute
in the
macro

one

of the of

major advantages
smaller

of the

model

compared to the
clear

micro model

a context

comparison of design parameters. Another generally

were

advantage of
no

model is its

computational

effort.

Deliberately,

direct CPU

time
that

comparisons etc.,
reason.

performed

since the models

are

essentiaUy too different for

rapid Computer perhaps
model smaUer and

Another

important aspect of

this matter is the very of the

a
macro

development,
less

which will make this

advantage

important in the future. However, presendy

to

reasonable micro model wl

typically
was

involve at least twice stl very noticeable

three times the number of degrees of freedom, and this fact Solution of the nonlinear time

during the

history analyses of the multi-

storey buding.

The micro model,

as

developed in this report, mainly fulfilled the purpose of a compa

model. This

rison model for the

macro

comparison should be compcated problems.

history comparisons

seen more as an

existing
use a

tool to work with and to be avaable for

In this report its final

macro

mainly
case

suffices for
no

some

nonlinear time

with the

model for

where

experimental

data is available. Thus, when each model fulfilled its

tests

comparison with experimental static

reastic. The micro model
same extent as

satisfactory

and showed agreement in in this

case

dynamic

analyses, it appears likely that their respective behaviour also

may be

quite
die

as

presented

in this

study does not eontain

new

development

to

die

macro

model, which is in accordance with its primary

use as a

comparison

and reference model. Most of the

development steps
may be

may in their
a

essence

be

found in the hterature. However, in the present form it constitutes

clear and into

an

logical

development,
of

and shows how certain

phenomena
are

simplified

efficient
areas

model. Particular
a

strengths of this model

the

possibity to look in deta at local

as

structural wa, and examine local of reinforcement. These local

damage, such

cracking,
may be

crack directions, and mueh

more

yielding

damage analyses

performed

accurately with die micro model than with the macro model.
die micro model may be connected with
a

In the present

implementation
Zrich
are

graphic post-processing Software, developed

by
in

Mr. Thomas G. Wenk of the Institute of Structural

Engineering of the ETH in

Cooperation with
in

the author. Some of the initial

capabities of this graphic

Software

seen

figures 5.53b-d, currentiy comprising magnified deformed shape and crack direc
yielding
of reinforcement in different directions and
are to
a

tions at any time increment. The

more

refined crack indication

revealing closure and reopening,

such

be

implemented.
macro

These

damage analysis capabities make the micro model clearly superior to the analysis cases,
as

model in certain

die

analysis

of damaged

existing wall

structures.

148

CHAPTER FIVE

Another case where the micro model is

superior is

when die wall geometry is

irregulr,
geometrie

exhibiting openings for Windows and doors,

discretization may the be

which is

quite often

the

case.

The

performed readily with

the micro model, since it

to use.

uses

membrane

elements, whereas the macro model may be almost impossible

Some clear drawbacks with the micro model
are

its

relatively complex Constitution.

It
as

basically consists
described in

of

modulus matrix, which is made up of different contributions matrix forms

a

chapter five. This modulus

a

part of die stiffness matrix formu

In

lation, together with

displacement-strain transformation matrix.

due to this

addition,

local-

global transformation matrix is generaUy needed for die orthotropic

upon

concrete

behaviour

cracking.

The

major drawbacks
it is its

complexity

Ue

partly

in the code and

model
or

development:
a

generaUy difficult for other programmers to continue developing

to

improve

model due
to

complexity. Secondly,

these models need

a more

refined

discretization
and after that

work we.
more

Thereby more effort is needed to

numerical effort is needed

generale die element mesh,

clearly

during

the Solution. This is,

as

already mentioned above, especially noticeable

solutions. With

for extensive

dynamic time history

rapidly increasing Computing power, and thereby decreasing CPU time, computational effort of the micro model wl become less powerful mesh generators
the
and

the drawback ofthe extensive

important

in the future. More

post-processing

Software

wl also reduee the drawbacks of the micro model. As for the

cross

sectional

quantities avaable for

by

macro

model, these may

not

be

produced readily using

the micro model. Some

possibities theoreticaUy exist of creating

the summation of the proper reaction of all

e.g. base shear and base

overturning

moment

forces, and handling by

cross

graphical post-processor. However, the ready avaabity

at any

sectional
not

quantities

level of the wa

as

obtained

by

the

macro

model may
also

generaUy
provide
For

be avaable. However,

special post processing

in

could

theoretically

that. the

checking
macro

dynamic Performance
presendy offers the

design

process

involving multi-storey
The micro model
some

was, the
has still

model

most

convenient and efficient tool, which also

some

development potential

for three dimensional

on

analysis.

currendy represents a more compcated tool, analyses,

of the
as

the other hand suitable for

special

mentioned above, and possesses considerable

development potential in view

rapidly growing Computing power.

numerical

By these comparisons, the chapter on

of
a

examples is concluded.

A discussion

number of design

problems within

the framework of the

capacity design method,

in

regard to numerical results, wl be presented in the subsequent chapter.

149

CHAPTER SIX

CAPACITY

DESIGN

CONSIDERATIONS

6.1

Introduction

Among

die efforts to achieve effective resistance

against seismic

actions

on

reinforced

concrete

buildings,

the

capacity design

method

[PBM90] has become particularly

successful. The method is

systematic and oriented towards practising engineers. Relative

are

ly comprehensive
A

recommendations for the

provided for the structural detailing.

verify
the seismic

major reason

development of the numerical models described in the previous

tool with which
one can

chapters,
structure

is the need for

Performance of a

and relate it to the

Performance expectations
in the

set out at

the

design phase.

Based
with

on

the numerical models the

presented

previous chapters, this chapter deals

structures.

questions regarding
on

Performance

of

capacity designed

The interest

focuses

multi-storey buding

structures in

which the lateral resistance

against seismic
as

action is and

provided by

structural was. These was should act

essentiaUy

cantilevers

preferably the nonlinear behaviour should be mainly of a flexural nature.

is not described here, but parts of it
a
or

The

capacity
as

design procedure

are

briefly

introduced of the

far

as

relevant in the context of the present studies. For

complete presentation
[PP92].

capacity

design

method the reader is referred to

[PBM90]

Since the fundamental idea of the

capacity designed structural wa is the definition of a

plastic hinge
which may

zone

located at the base of the wa, and the

protection of the rest of the wall

magnitude
of flexural

against yielding, strength,

it is of

major importance
flexural

to assess

the

strength

develop in

the

plastic hinge. During

overstrength possible
a

severe

seismic action the moment at this

referred to

as

moment, is transferred to the

remaining
The

elastic part of die wall. In order to

safeguard against yielding

flexural

in this part, it is thus of of the

importance
elastic

to

know the level of

overstrength
a

plastic hinge.

region must be designed in such

zone,

way that at

bending moment which

overstrength

acts in the

plastic hinge
Another

corresponding
reason

to the flexural

overstrength,

it must stl not is its effect

yield.
the

important

for

knowing

the flexural
zone

on

magnitude

of the shear force in the

plastic hinge

and in the elastic

region [PBM90].

150

CHAPTER SDC

For

chosen

global displacement ductility level,

to

and

selected wall geometry, it is

furthermore necessary

estimate the level of

curvature

ductility demand which

wl
zone

develop
is based

in the
on

plastic hinge zone. This is due

to the

fact that proper detaing in this

correctiy estimated deformations.

the curvature

FoUowing

ductity demand

brief
a

investigation is carried

out

of the The

energy distribution in wall that

develops during

nonlinear time

history analysis.

results of this energy balance wl demands

help to explain

the different

dynamic curvature ductity

compared to static calculations.

we

Then,

deal with the

question of the distribution of flexural strength demand during design. The curtailment of the flexural reinforcing forming of plastic hinges
at

seismic action, and the proper flexural

is
an

important aspect
one

in order to avoid the

other locations

than the intended

at

die wa base.
on

Lasdy,
time

the

dynamic base shear demand is discussed, based

results from nonlinear

history analyses. Comparisons are made to recommendations in the design xerature.

The five aspects introduced here: flexural

overstrength,

curvature

ductity demand,
dynamic
shear

energy balance, the distribution of dynamic flexural

strength demand,

and

demand wl be treated in the

following five sections of this chapter.

6.2 Flexural

Overstrength
of structural wa,

In die

plastic hinge

zone

large deformations

may

develop during
at

seismic action. These deformations may be

considerably larger than

0.35% strain
at

those

which the

design strength
tensile
material

is determined,

(usuay

at

compressive edge).
reserves

At the of the

edge strain
are

of several percent may

to as

develop. Thereby, strength

mobised, referred

material

overstrength. Consequeny, section forces

wl

develop, which are correspondingly larger than the section design strengths.

In the

capacity design terature [PBM90], suggestions

for sectional

overstrengths
on

are

presented to be used for structural

nonlinear time

was and for beams in frames, based

material

over

strength and assumptions regarding the section deformation mainly for an R/C beam.

For

history analysis

which attempts to achieve

reastic Simulation of the

structural behaviour under seismic action, it is

One of the most

important to catibrate the input parameters.

which accounts for the behaviour upon

important parameters

is the

one

yielding

and thereby includes die

overstrength.

CAPACITY DESIGN CONSIDERATIONS

151

6D20

Figure

6.1 Cross section

of structural wall, restricted ductility design

6D12

D10/250

Sym

.Afra.3 U.XPA &i:?<i .xm^^r&^^am^s*^.

ifcjo?4

Lw/2

3000

Figure 6.2 Cross section of structural wall, fll ductility design

For this reason,

we

perform

here

small numerical

study

of the

overstrength as
to

it may

develop

for relevant wa

cross

sections. The numerical

procedure designed
introduced in

compute
three in
model is

flexural behaviour of R/C

cross

sections which

was

chapter
macro

connection with die cabration of the nonlinear flexural behaviour of the

used here. The

same

material models,

including the concrete behaviour by

Kent and Park

and the idealised bilinear steel behaviour shown in

at

chapter three,

wl be used. The strain

which the concrete

compressive strength is reached, however, is modified from 0.002

0.003 in order
to

(as in equation 3.28)

wall
cross

to

model the concrete used

more

realisticaUy.
as a

The

section from the

The
cross

eight-storey buding
and 6.2. The

of chapter five is used

calculation

example.

sections for the restricted and for the fu

in

displacement ductity
was

design
Ec

are

given again
fc'

figure 6.1.

foUowing

set

of input data

used:

33000 MPa,

19.5 MPa, Es

210000 MPa,

fy

460 MPa,

soh

0.0054 (confined concrete), 50

0.0034 (unconfined concrete)

Strain

hardening stiffness

for steel: 0.7 % of Es

Strain at concrete

compressive strength: 0.003

Axial force: 4.15 MN

Number of fibres: 106

Equibrium

tolerance: 0.05 MN

152

CHAPTER SDC

Three

computations

were

performed.
strain

The first

one

utised the

design

values for

strength
mean

as

stated above, but

no

hardening.

The second
was

computation used effective

values for steel, i.e. the

yield strength

for the steel

adjusted tofy

550 MPa,

and in addition die strain

concrete

hardening was taken as 0.7

% Es. For the third

computation, the
mean

compressive strength was additionay adjusted to an effective

a

value of 30

MPa. Thus towards relations

a

gradual

move was
as

performed
as

from the

typical design calculation values

moment versus curvature

behaviour
shown in

realistic

possible. The resulting

are

figures

6.3 to 6.6.

Figures 6.3 and 6.4

6.6
are

are

based

on

restricted

displacement ductity figure 6.3

and

figures

6.5 and

based
was

on

fll

displacement ductity.

In

the effective normal force of

4.15 MN

included in the

computation
seen

and thus contributes

an

significandy

to

the inter from the and

nal moment. From

figure

6.3 it is
mean

that

increase of flexural

strength

design values
effective
seven mean

to

effective

values for

yield strength is around 10 percent,

value for the concrete

compressive strength gives additionay

overstrength of

around

percent
a

Thus
concrete
as

total

cross

sectional flexural

17 percent is obtained, where die

we

stl accounts for almost half of that amount Without the normal force,
at

obtain

expected roughly 20 percent overstrength

few percent

smaU curvatures when

using

effective

mean

yield strength (corresponding to the relation between die yield strengths: 550/460),
more at

and

higher curvatures due

one to two

to

the strain

hardening.

The concrete

only contributes with

that with

an

additional

percent,

indicating

that the concrete

strength has a smaUer effect when the

zero

normal force is removed.

Figure 6.4 shows further

compared
to

normal force the

moment

dropped

over

40 %

the

curves

including

the normal force of 4.15 MN.

For the fu
mean

ductility design figure 6.5

shows the

case

with normal force. Effective

mean concrete

yield strength overstrength

adds around 7 percent, and effective

strength
simar

additionay around eight percent

flexural

Thus the concrete accounts for over the half of the total

of 15 percent

Figure 6.6 shows

that for

zero

normal force

behaviour as for the restricted ductity

design is found.

We obtain here e.g.

relatively
are

low

overstrength values compared with suggestions from

use

[PBM90] which

intended for

with beams and was based

on

calculations of
are

beams. In these sectional

suggestions values

in the rnge of 18 to 41 % percent

given

for

overstrength, based upon material tense overstrength (for 2% and 4% tense

strain, respectively) and considering the effect of the steel only.

CAPACITY DESIGN CONSIDERATIONS

153

25

-i

>

r"

'

'

'

20

"E
15
-t-t

il
it

E
o

10

if il il

fy fy
fy
i

460, ay
550, ay

0.0 %,
0.7 0.7

fc

=19.5
=

%, f. %, fc

19.5
30.0
-

550, ay

0.000

0.002

0.004 Curvature

0.006

0.008

(rad)
design

Figure 63

Moment vs. curvature relation incl. normal force, restricted ductility

25

,.

..j

,..,-.......

,.,,,..

20

fy
fy
-

460, ay
550, ay 550, ay

0.0

%, fe =19.5

?
z

0.7 %, 0.7

fc

19.5
30.0

2
4->

15

fy

%, fc

c o

mTm-

F
o
->

10

/
i
...

...

..

0.000

0.002

0.004
Curvature

0.006

0.008

(rad)
zero

Figure design

6.4 Moment

vs. curvature

relation with

normal force, restricted

ductility

It was here shown that

mainly due

to

the effect of the normal force the relative effect of

same

the steel is mueh smaUer, and in addition that the effect of concrete is of the that of the steel. Thus, for was with axial force the total flexural
than

order

as

overstrength is

lower
the

anticipated

for beams, and the concrete

accounts

for

signifieant part of

overstrength.

154

CHAPTER SIX

25

-i

->

-i

r-

E
z

fy fy

460, ay
550, oty 550, ay

0.0

%, fc

19.5 19.5
30.0

0.7 %, 0.7

fc

fy
0

%, fc

0.000

0.002

0.004
Curvature

0.006

0.008

(rad)

Figure 65

Moment vs. curvature relation including normalforce, fll ductility design

25

-ii

"

'

20

fy fy

460, ay 550, ay
550, ay

0.0

%, fe

19.5 19.5
30.0

?
z

0.7 %,
0.7

fe

2
-->

15

fy

%, fe

c tu

6
2

10

"

^m.

/j"

~~

/
0
j

jiiii1_.

_ii

ii...

0.000

0.002

0.004
Curvature

0.006

0.008

(rad)
zero

Figure 6.6

Moment vs. curvature relation with

normal force, fll ductility design

The conclusions that may be drawn from the numerical smdies sections
can

on

die two wa

cross

be stated

as

foUows.
no or

For beams and for structural was with

smaU normal force there is

usuay suffi
steel

eient concrete

area to

accommodate the

compressive force without excessive softening or

suggested in [PBM90].

crushing of die concrete. This allows for large tense strains and forces. The large
strains
account

for overstrength behaviour

as

CAPACITY DESIGN CONSIDERATIONS

155

However, for was with considerable normal force, the compressive strain exceeds
the

point

of concrete

strength

mueh sooner, and

more.

maximum

compressive

force has

developed
strains

and cannot be exceeded any

The neutral axis has reached its

position

farthest from the centroid, and wl tend to

move

back towards the centroid. The tense

developed

are

only

as

large

as

needed for the steel forces to balance die

account

bending

part of the

compressive force.

These limited steel strains

for

less dominant

effect of steel in the total

case, accounts for a

overstrength.

The concrete, which is the

limiting

factor in this

relatively large part ofthe total overstrength. overstrength plays emphasised that
which

There

are
as a

several aspects where the flexural

well
as

an

important role.

In

[PBM90]

many other references it is

are

the shear forces which

develop in
moments

structural wa

dependent on the moments

overstrength

develop.

And if larger

develop (flexural overstrength) this

the flexural

wi have die consequence of

larger shear

forces.

Accordingly,

is considered in the

capacity design
wl be
State

procedure in the calculation of the demand of shear force. When it was found in the above
numerical studies that for was with smaUer than

high normal force,

the flexural

overstrength

anticipated with simpler hand calculations according [PBM90], one may Suggestion is a conservative estimate, i.e.
obtain
use
an

that the hand calculation

on

the safe side. It is,

however, of importance

to

estimate of how conservative the

Suggestion is, Perhaps

and

that may be achieved with

of the above used Computer Simulation.

more

economical and reastic

suggestions may rise from die more exact procedure

used above.

The next aspect where the flexural

overstrength is important is plastic hinge

in the upper part of die

die flexural

design of
region
be

the wa

cross

section in the

region

where the

ends and the elastic

starts. The internal moment

developed

plastic hinge wl

transferred to lowest part of the elastic

region. Since this region should remain elastic

as

it is

important to have
be
seen

as

good an estimate
chapter,

possible of what moment is transferred. As wl place additional longitudinal

this

later in this

it may be necessary to

reinforcing bars in this region in order to fulf

requirement.

6.3 Local and Global

Ductility

Demand

In order to

adequately deta
zone

plastic hinge

zone

it is necessary to

assess

the

amount

of

deformation this
zone

may

undergo.

The curvature

ductity demand in
in two aspects.

the

plastic hinge
this
measure

of
an

structural wa is

particularly important

Firsdy,

gives

estimate of the deformations for which the reinforcement

detailing must be per

to

formed.

Secondly, it wl

influence the amount of shear force the wa has

withstand.

156

CHAPTER SIX

"9-1-- 20
ii

"^

4.

18 16 14 12

5
Jmt

-o
to

o> c

E
:3

10
8 6
4
L

/
"'"

QJ

,.,,^*
I

o
i

2h
0 0
2
4

10

12

14

16

geometrische

Wdndschlonkheit

h^/l^
ratio and

Figure

6.7 Curvature

ductility demand

as

function of slenderness

global

displacement ductility level [PBM90]

Both

experimental

and

analytical investigations [PBM90]

ductility
demand for
on a

have been

performed

in

order to suggest the curvature

structural wa. The

experimental
Analytical
a

investigations comprised
increasing

static model tests

wa members and have shown that for

demand wi increase.

wall slenderness, the curvature

ductility
at

studies of a cantilever with elastic behaviour except

die

hinge,

and

subjected to

static

point load

at

the free end

give
a

similar results,
area

see

figure 6.7.
rather than

For each chosen

a

global
area

displacement ductility
Covers

level

shaded

is

given

Une. The shaded The upper

as

different

assumptions concerning
represents

the

length

of the

plastic hinge.

edge

of the shaded

areas

yielding assumed

over a

hinge length

taken

half the

horizontal wall length, whereas the lower edge

to

assumes

yielding over a length taken equal analytical smdy

in

the horizontal wall

length.

It should also be mentioned that in the

[PBM90]

it

was

assumed that uniform

yielding takes place

over

this assumed

plastic

hinge

area, and

totally elastic behaviour is assumed over the rest ofthe wa. analysis considers
modes
are

It is clear that the static

the

dynamic

behaviour of the first mode

only,

and the effects of

higher

disregarded.

With the numerical model of

chapter three a dynamic analysis series is performed in order to investigate die dynamic
curvature

ductity demand.

CAPACITY DESIGN CONSIDERATIONS

157

Curvature

Ductility
i
i

Demand
N

K
o
N N

\lA

5, Cty 5, Oy

0.01

8
CS

12 11

uA

0.03
0.01
0.03

II

HA=3, Oy
liA
=

10 9 8
7

3, cty

6
5
4

-m

3
2
1
I I

I.

aspect ratio

10 11 12

Figure 6.8 Dynamic

curvature

ductility demand,

as

function ofwall aspect ratio, and

global displacement ductility

The

same

eight-storey

structure as

used in

chapter five
=

wl here be used
a

as

reference.

The aspect ratio of the reference wa is 32m / 6m the aspect ratio

was

5.33. For

number of further was, the

adjusted by changing adjusted

to

die wall

length. The reinforcement of

strength. Changed

comparison

walls

was

give

proper flexural

fundamental

frequency of Vibration was considered for a comparison walls giving different demands
from the static
reinforcement

equivalent

force calculation. For the

more

slender walls minimum

requirements was mosy governing.

three
macro

The

plastic hinge was discretised into

although stricdy

elements, with

total

length of 6m

(taken from the reference waU), and this arrangement was kept for mesh consistency for
aU tested walls,
not

entirely

correct
a

for the two most slender walls.

However, this

measure

is rather to be

regarded as

"construction"

length

with

detaing
wl be

allowing

for

major yielding,
area

whereas the extension of the effective the numerical

yielding

determined within this numerical

by

analysis.

A clear difference between the

analyses
the

and the

analytical analyses of figure 6.7

made

is thus that for the It

was

analytical entirely

analyses a clear division

only within
elastic. In the numerical

was

concerning the yielding.

The rest of the wall

assumed uniform and

plastic hinge length.

was

assumed to be

analyses, however, yielding proceeded along

it to the extent that
were was

the

plastic hinge
the

height and possibly above history analyses.

No

determined

during the

nonnear time

particular attempts

made to suppress

yielding right above

158

CHAPTER SDC

plastic hinge,
wa

other than that the flexural

design

was

entirely performed according to

the

recommendations in [PBM90] with curtament of the flexural bars

along the height of the

section 5.4.2 of
no

according

to

the recommended

linearly decreasing line discussed in

to

[PBM90] (shown also in figure 6.13 in this report) according

should be aowed to take place.
The 10 second SIA

which

yielding

compatible ground

motion
were

input

as

discussed in
under the

chapter five

was as

used, and nonlinear time history analyses

discussed in

performed

same

premises

chapter five.

Figure 6.8 shows the dynamic curvature ductity demand obtained from the numerical comparisons. The values
stiffness ratios cty, for the
section 3.5
were are

taken from the element dosest to the wall base. Yield

outer

flexural

spring of the

macro

elements,

as

discussed in

taken

as

0.01 and 0.03. The

shown in

higher yield stiffness gives lower curvature

In

ductility demand,
two

as was

chapter five, section, 5.4.2.

is thus

figure 6.8, each

area

of the

chosen

design ductity levels

represented by

shaded

where, for each

aspect ratio, the upper obtained

stiffness ratio.
In the

curvature

ductity demand arises

from the lower

yield
A

figure,
to

the first natural

frequency

of the calculated was is also for

displayed.

tendency figure.
at

lower curvature

ductity demand
to

higher aspect

ratio is

seen

from the and

The

tendency appears

be

more

pronouneed for higher design ductity level

extracted from nonlinear time

higher aspect ratios.

This difference

compared to figure 6.7 is mainly explained from history analyses,

the upper

the fact that the results in where direct time

figure 6.8
was

are

integration

employed, thus containing a higher modes. For more

more

flexible walls, i.e. with storeys where

high aspect ratios,

and
some

energy is takes

dissipated
place,

over

major cracking
at

yielding

and the relative

concentration of rotation
ofthe

the

plastic hinge
load for floor

is decreased.

Secondly, only the lower third

total of 4.15 MN

plastic hinge is regarded, giving higher curvature ductity values.

to

Normal forces due

wa
cross

gravity ground

eight

floors

giving

on

the
in

section

at

the

were

employed throughout

the

analyses shown

figure 6.8. Since the results of static calculations in figure 6.7 apparendy did not consider
the effect of normal force, the smaUer gravity loads, down to
same
zero

numerical

examples of figure 6.8 were repeated with

load, without changing the results significany.

It should further be noted that the shown

global displacement ductity level in figure

not

6.8

only pertains

to

the

design phase,
easy,

i.e. it is
as

measured

during

the time

history
the

analysis,
reason

and cannot be defined

discussed in section 5.4.2.

Briefly stated,

is that, due to influence of higher modes, the roof level may not be

displaced in the

CAPACITY DESIGN CONSIDERATIONS

159

'^roof

Figure 6.9

Global

displacement behaviour for monotonic static loading,

used to obtain

displacement ductility from time history analysis

direction the curvature at the wa base. This

same

as

gives possibities displacement

for
at

positive

or

negative global displacement ductity

when

if based upon roof

the instance
may

yielding begins
manner.

at

the wa base.

EssentiaUy, infinite global ductity values

result in this

One

possibility

of

obtaining

reasonable

global displacement ductity

as

based upon

results from the time

history analysis,

is described

foows. The maximum roof level

is extracted. Then,
a

displacement
static

obtained from the time

history analysis

monotonic

analysis is performed on the same wa, using an inverted triangulr equivalent static

force pattem, roof level static

set

apped incrementaUy.
reached

The increments

are

added until

staticay

the

same

displacement is

as was

obtained for the time

history analysis.

In the

analysis
6.9.

the onset of yielding at the waU base is

kept as
The

reference, which is then

is illustrated in

in relation to the maximum roof level

displacement

principle

figure

The

resulting global displacement ductility,

from the time

based

on

the maximum roof level

as

displacement

history analysis

wl here be referred to
as

the

dynamic

displacement ductity, ^A*".

dy*

It wl thus be obtained

(6.1)

160

CHAPTER SIX

The assumed
not

global displacement ductility

found

values

at

the

design phase, wl generally

wall the

be reached

exacy by this

method. For the

were

eight-storey

global

ductiUties

obtained using the above method

For assumed restricted

to

be:

ductility (Pa
=

3):

jua*r"
=

2.9

For assumed fll

ductility O^a
the

5):

fiA*~

5.3

In this

example

correspondence

between the

ductity values

obtained

by

this

method

to

the chosen

global displacement ductilities is thus relatively good.

A number of

other methods of determining discussed in

global displacement ductity from time history analyses are

[Wenk93].

6.4

Energy Dissipation
comparison
between walls of different aspect ratios
It
was

In section 6.3

was

made

regarding

the curvature

ductility demand.
from static

found that for

high aspect ratios,

zone

i.e. for flexible

be smaUer

was, the
than the

curvature

ductility demand

of the

plastic hinge

mmed

out to

expected
smdy

analysis of elasto-plastic cantilevers. This fact is explained by dissipation during

an answer to

assumption that more energy is dissipated over die upper storeys for flexible was.
here the energy

We

time

history analysis

of two of the pre

some

viously used was,

terms are

in order to get
as

these

assumptions. Thereby

energy

introduced
are two

foUows.

There

types of strain energy. The first is the elastic strain energy

E.t,

(recoverable strain) which for an element may be written

ESe=~uTk.u
where the is the element

(6.2)

displacement

vector, and k. is the elastic element stiffness

matrix.
The second strain energy type is die inelastic strain energy

E^ (irrecoverable strain)

also known

as

the

hysteretic energy,

which may be written

,*

1
=

~u

kuu

(6.3)

where

ku

is the inelastic element stiffness matrix.

CAPACITY DESIGN CONSIDERATIONS

161

140 120

-iir-

Totol transmitted energy

TotGl

Plostic
E

dissipation hinge dissipation

80

Figure 6.10 Energy dissipation of eight-storey wall, aspect ratio ra

Two
more

533

energy types

are

defined

as

the viscous energy,

Ev,

which may be obtained

by integrating the viscous effect Py

1
=

*r

cu

(6.4)

E.=\Pmdt
where is the

(6.5)

velocity vector and c is

the

damping matrix,

and

lasdy the kinetic energy

Ek, defined

as

Et
*

mu

(6.6)

where

is the

mass

matrix. The total energy

as

E,

transmitted

to a structure

by

ground

motion, may then be defined

Et

Et.

Etl. + E9

Ek

(6.7)

We

are

here
two

especially

interested in

dissipated

energy, and of the above energy

ex

pressions,

contribute to the energy

dissipation, namely

the inelastic energy, and the

viscous energy,

together accounting for the dissipated energy Ed

162

CHAPTER SIX

140

-'

'

'

'

'

'

'

'

'

'

'

'

'

'

120
--

Total transmitted energy

Tl

Total

100

Plastic

/ 1 dissipation hinge dissipation J \

f^A

r-Y1

E
ZT

80

cn
l_

60

j\)
.PA/W
r

^r-'~~
''
+-*
~~

40

\ f

20 0
-_*-LS-TTi

J
""*

,y~
1 \ 1

_i.

I_l

10

Time

[s]
ra
=

Figure 6.11 Energy dissipation ofeight-storey wall, aspect ratio

8.0

E,=E.
d
sie

+E
'-'-,

(6.8)

We

mainly

focus here
rest

on

how the

dissipated energy
as

is divided between the

plastic
5.33.

hinge
For
was

area

and die

of the wall, referred to five

was

the elastic
an

region.

For this purpose the

=

eight-storey

wall of

chapter

selected, with

aspect ratio of 32.0/6.0

=

comparison

purposes, the flexible wall with aspect ratio of 32.0/4.0

8.0, which
for
a

used in section 6.3,

figure 6.8,

is chosen. Both selected was

=

were

designed
here

restricted

global displacement ductility Qj.a

amount

3).

In order to limit the

most

of Output data, and for

energy,

clarity,

we

display

only

the
a

relevant data. The total

as

dissipated

consisting of the contributions from

time. The contribution of the

inelastic strain

well

as

the

slight viscous Rayleigh damping of around two percent,

versus

is

displayed

as

accumulated wiU be

dissipated energy

plastic

hinge elements Figure 6.10

displayed separately

so

that its relation to the total

dissipated

energy may be estimated. shows the

dissipated energy for the by

wall with aspect ratio 5.33. The total

energy transmitted to the structure is shown

the solid ne. The difference between the

total energy and the total energy.

dissipated energy is made up

same

of elastic strain energy and kinetic 8.0

Figure

6.11 shows the

quantities for the wall with aspect ratio equal to

CAPACITY DESIGN CONSIDERATIONS

163

It is

seen

that for the slender wall the amount of total

dissipated energy

is increased,

and this in absolute terms and in relation to the total energy. This may be

partly explained

by

lesser elastic behaviour overa, and

by lower first periods. plastic hinge

accounts

Secondly, dissipated

one

may observe that the

for clearly less of the total

energy in the flexible wa. The relative increase in

to more

dissipation

of the upper

storeys is related

flexural

cracking,

and partial

yielding.
die
to

Based total

on

the stiffer

waU, the decrease of the contribution of die plastic hinge

to

dissipated

energy may be estimated from

a

figures
a

6.10 and 6.11

as

from 35 %

around 24 %, which is
It
was

relative decrease of roughly

third

possible

to track the

increased nonnear behaviour of the upper storeys

by

means

of the

macro

model. Some results in this

regard were already presented in chapter

five. However, in the

more

subsequent section

of this

chapter, this behaviour wl be dealt with

in deta.

Suggestions on how to encounter the problems associated with outspoken

nonlinear behaviour of the upper storeys wl then also be discussed.

6.5 Flexural

Strength

6.5.1

Impcation of numerical results plastic hinge zone is detaed, the question of the flexural strength for die rest of assumptions
of
an

Once the

the structural wall rises. Based upon

equivalent

static force,

Suggestion has been made [PBM90] concerning the distribution of flexural strength from
the

plastic hinge

zone to

the top of the wa. This

Suggestion is visuased in figure 6.12.

entire

The essential features of this

Suggestion

are as

foUows:

The flexural

strength is kept

constant for the

plastic hinge

zone,

stretching
taken
as

length Lp upwards
horizontal
of the flexural
zone to

from the base of the wa. Values for

a

Lp

are

usuaUy
a

the

length of the wa Ly,, or strength

is until

fraction thereof. Above this zone,

linear decrease

suggested stretching
or
a

from the upper end of the

is reached which

plastic hinge
to

the top of the wall,

flexural

strength

corresponds

minimum reinforcement

requirements.

It is further assumed that the distribution of the effective

bending moment acting over

static force

die wa

height has a shape simar to that of a cantever subjected to a lateral

with inverted triangulr distribution.

164

CHAPTER SIX

Biegewid erstand
der Mindesfbewehrung

Erforderlicher

Biegewiderstand

Momente

infolge

der Ersatzkr'ff e

--V)7rV7777?T
N

Biegewidersfond
am

Wandfuss

Figure 6.12
With the

Distribution

offlexural strength, proposed by [PBM90]

ability

to

perform

nonlinear time

history analysis, using

moment

the wall elements

developed in chapter three,

wall
wall used in

we now

examine the effective

distribution

acting on a
the

subjected to ground motion.

chapter five.

The

bending moment distribution is

regarded
as

shown here for the

This wall may be

to

relatively

stiff. In

figure 6.13

distribution for this waU is shown slender wall of the

same

the left, and in addition the distribution for

a more

height

shown to the
or

right.

The fundamental natural

frequency

without any nonstructural elements

frame
were

stiffening

is 0.67 Hz for the stiff wall and

0.40 Hz for the flexible wa. Both was

ratios
are

designed for restricted ductity. The aspect design quantities

are

5.33 and 8.0

respectively.
as

The most relevant flexural

summarized for both walls

follows.

ra
Fundamental

5.33

ra

8.Q

frequency (Abaqus) equivalent

0.67 Hz

0.40 Hz

Moment at wa base from

Static force calculation ME

Moment demand

16.1 MNm 19.3 MNm

19.4 MNm

8.9 MNm 10.6 MNm

11.1 MNm

Af,-

y^ M-

Design strength Mr

CAPACITY DESIGN CONSIDERATIONS

165

M
-25-20-15-10-5 0 5 10 15 20 25
/x__
.

/
-25-20-15-10-5 0 5 10 15 20 25

f1=0.67Hz

(MN111)

fi=0.40Hz

(MNm)

Figure 6.13 Effective moment distributions extractedfrom nonlinear time history analysis
for eight-storey wall [BWL92b]
more

with aspect ratio

of533 andf\=0.67 Hz (left),

andfor

slender eight-storey wall with aspect ratio

of8.0 andfi=0.40 Hz (right)

shown to the

It may be

clearly seen that for the more flexible wa (moment distribution

modes have
more

right),

the

higher

influence, i.e. higher

out

moments are

obtained at mid

height of the

structure.
as

It should

again be pomted
moment

that the

moments at a

particular
history
same

storey are extracted

die maximum

reached at that storey during the time

are

analysis, and thus

time.

the values for different storeys

generaUy

not

obtained at the

In order to further

explain

the

tendency shown

in

figure 6.13

we

examine

eigen

frequencies
From

of die two walls,

seen

in the response spectrum of the

ground motion input,

smaU

displayed in figure 6.14.

figure 6.14 it is
to

seen

that the first mode of the flexible wa has

spectral
be im

value

compared

the stiff wa.

Relatively

seen, the

higher modes wl therefore

towards the left in

portant for the behaviour of the flexible wa. As the nonlinearities occur during the time

history analysis,

a the

eigenfrequencies
more

wl

generaUy

move

figure

6.14, i.e. the system becomes

flexible when

damage oceurs.

166

CHAPTER SIX

Flexible wall

|fl

{f2 j

f3

Stiff wall

jJ4%g
16% g

0.2

0.5

1.0 2.0

5.0

10

100

f[Hz]

Figure 6.14 Eigenfrequencies of structural walls shown above

input ground motion.
It is clear that the

response spectrum for

shape

of the

ground motion response spectrum plays

wl be

an

important

role, and thus the relation between the structures' eigenfrequencies, and the shape of the

input spectrum wi determine to what extent the higher modes

As
a

important or not
frequencies
moment

consequence, it is possible

to State

that the relation between the structural

and the
over

shape of the input spectrum wl determine the distribution of bending height of the
structure.

the

The

principle may be iUustrated as in figure 6.15.

the above

It is thus clear that the flexural

strength distribution should be dependent on

a

relationship.

It is, however, difficult to obtain

general relationship including ground

motion, structural eigenfrequencies, and

Stiff wall relative
to

moment

distribution.

Flexible wall relative

to

ground motion

ground motion

Hl

Moment distribution

largely of first mode

consisting

Moment distribution with

higher modes

Figure 6.15 Influence on moment distribution of relation between wall eigenfrequencies

and ground motion

CAPACITY DESIGN CONSIDERATIONS

167

Therefore, it may be reasonable

to

estabsh

strength distribution criteria

SIA. A

for

given

ground motion,
upon the above

e.g. the

design ground motion according to the

Solution to the flexural

Suggestion, based

findings, for a

strength distribution is described in

the

foUowing.

6.5.2

Suggested flexural strength design

how to avoid flexural

Suggestion wl here be presented thereby fulfil

one

as

yielding in the elastic region,

The

and
on

of the basic

goals

of the

capacity design.

Suggestion

is based

results obtained

by numerical analyses with the macro model developed in this report. strength was reduced over the height of a multithe

These results indicated that if the flexural

storey structural wa,

yielding would generaUy occur over the region intended to remain

design Suggestion
on

elastic. We wl here base the

moment at

assumption that the overstrength

at

the lowermost end of the

plastic hinge (i.e.

die waU base) may

develop

as

we

at

the uppermost end of die

plastic hinge,
the

and transfer to the

immediately bordering modeUing

in this

elastic

region.

This

simplification

is based

on

findings

from the numerical

report, which show that although the

the
moment

curvature

drops over the height of the plastic hinge,

due to the nonnear moment curvature

only decreases sghdy. This is

relationship, discussed earer in this chapter.

Quatatively, hinge
of a shown in
the distribution of curvature and moment

over

the

height of the plastic

may be iUustrated
as

multi-storey wa, discretised into three macro elements,

In

figure 6.16.

figure 6.16a,

distribution of curvature
moment

along

the

plastic hinge

height is shown,
moment curvature

and in

figure

6.16b the

distribution. The

corresponding

relation is

given in figure 6.16c. The numerical results from the macro

some

model smdies indicated that

yielding always occurred over the uppermost element although

the curvature of the upper end of the
one at

during time history analysis.

This has die consequence that

plastic
only
a

hinge

may

only

be about half or

third of the

the wa base, the moment is

little smaUer. This effect is iUustrated

clearly in figure 6.16c.

moment

Since the difference between

the moment at the wall base

M\ and the

of the uppermost element of the

to

plastic design

hinge M\\ is rather small

purposes,

it appears reasonable

neglect this difference

for

especially

since it is very difficult to estimate without extensive nonnear

dynamic analysis.
thus

The

assumption that M\ should be transferable

to

the elastic

region is

only sghy conservative.

168

CHAPTER SIX

Wa

height

height

03

Mt.

02 01
?
**

M2
M,
M

a) Curvature distribution

b)

moment

distribution

c) Moment versus

curvature

relation

Figure 6.16 Qualitative moment curvature distribution over plastic hinge for multi-storey
structural wall, discretised into three
macro

elements

Figure
known

6.17 shows

Suggestion
The

on

how to distribute flexural

strength

in order to

avoid undesired
to

yielding.

strength over the plastic hinge height is obtained according

and is denoted with Rp in the

capacity design principles [PBM90],

figure.

Above the plastic

hinge zone, the wall must behave elastically, i.e.

above the

no

major yielding is
which may

permitted. Directly

plastic hinge

zone

the flexural

overstrength

develop
The

in this

zone

is transferred to the elastic

region.

The elastic

region

constitutes the

rest

of the wall and has the

to

length Le

in the

figure.

strength in
This

the elastic

region bordering

the

plastic hinge zone

is denoted Re in the

figure.

strength must be such that it can take the overstrength from the plastic hinge

CAPACITY DESIGN CONSIDERATIONS

169

i,

i,

1 ' 1/

/4
h
IL
.

Lec

f-ec\

"A.
Flexible wall

Stiff wa

Figure

6.17

Suggested flexural strength distribution over height ofwall

i.e. its tense strain
exceed about 0.0025. This

zone

without

yielding,
an

must not

means

that there must be

A

increase of flexural reinforcement

right above the plastic hinge

plastic hinge
denoted

zone.

general way to State the required constant strength of the elastic region denoted Re in
6.17 is in
terms

figure

of the avaable

strength

in the

Rp

in

figure

6.17. We thus obtain

Re

k0Rp
is the

(6.9)

where

kg

overstrength factor for reinforcement steel which is usuay taken as

this
the

1.2.

Depending on what die effective bending moment distribution looks like,

must be

strength
findings

kept

constant over a

distance Lec, is

seen

in die
on

figure. According

to

earer in this

chapter,

this

length

dependent

the

eigenfrequencies of

the wa in
we

relation to the spectrum of the

may thus be able
must

ground motion.
find that the

For

given ground motion spectrum,

may be short for
a

generaUy
more

to

length Lec

stiff wa, but

be

longer die

flexible the wa becomes. In

shown to the left, and that for
a

figure 6.17
a

the

strength

distribution

for

typical stiff wa is

typical flexible
is

wa to the

right

Above the

height Lec,

linear decrease of flexural

strength

suggested, by curtailing

the vertical bars. If a minimum

strength due to minimum reinforcement i^uirements from

strength must be kept constant to
estimate of the wl be
a

the code is met, this minimum

the top of the wa.

It is necessary to estabsh
we

an

length

of constant flexural

strength.

As

have

already
Lec/Le.

seen, this

length

fraction ct. of the total elastic

length, i.e.
(6.10)

ac

170

CHAPTER SDL

Theoretically, this fraction will

values wl seldom be reached in
with the SIA

vary between

zero

and
a

one.

However, these
motion
a

extreme

practice. Assuming possible

and the

ground

input compatible
related to SIA that the modes

design spectrum,

it should be

to

estabsh

relation between die

length

of constant flexural

strength

eigenfrequencies (now being

a manner

spectrum).

It would be desirable to estabsh this relation in such

immediately higher
and since

than die fundamental mode will govem the result in addition to the

first mode. However, since this

our

procedure
a

would be

quite involved
relation

for

design purposes,
we

goal

is

to

estabsh

design guideline

which may be
our

ready usable
so

wl

attempt

to use

only the fundamental mode

and cabrate

that it agrees with

the results from nonlinear time

history analysis.
as an

Thus the desired relation would have

input

the fundamental

eigenfrequency
to be

of

the wa, and

a constant

as

Output the fraction of the elastic wa length which has

designed for
that

flexural

strength (corresponding
over

to

the

overstrength

in the

plastic hinge zone).

The effective moment distribution

the fraction of constant

the

height of the
to

wall in

figure 6.13 impes

strength
to

be

proportional
0.2 T\, i.e.

the inverse of the fundamental We thus have ctc

frequency,
to

i.e.

proportional
ct.

die fundamental
as

period T\.

proportional

T\. We here suggest

be taken

ac=0.27;

(6.11)

ae=0.2
h

(6.12)

It is clear that this

simple Suggestion
too

cannot be

regarded

as

generally

valid for

frequency regions
be used, e.g.
a

which have not been tested in this

study.

For very flexible walls the refined method should

Suggestion would give

high

ratios. For such

cases some more

direct inspection of the effective moment demand.

6.5.3 Numerical

example
will here be iUustrated
two

The flexural

design suggested above

by

numerical

example

which is also discussed in

[BL93] involving

walls with different fundamental

eigenfrequencies due to different aspect ratios.

with aspect ratios of 5.33 and 8.0
a

The two

eight-storey was of figure 6.13

were

respectively wiU

be used. Both was

designed for

global displacement ductility factor Pa equal to

Using expression (6.12)
for the
two

three.
we

walls of figure 6.13

obtain

CAPACITY DESIGN CONSIDERATIONS

171

Wall with aspect ratio 5.33: ac

Wall with aspect ratio 8.00:
ct.

0.2
0.2

1/0.67 1/0.40

0.3
0.5

Placing

these

strengths
as

over

the effective

moment

distributions

gives

reasonable

coverage of the demand,

=

Ulustrated in figure 6.18 for the stiff wall with aspect ratio ra

5.33. For the demand the maxima of negative and

positive effective moments

are

taken

from

figure 6.13. just schematically performed

die the top of the wa. A
no

The curtailment is here

to

simple
normal

curtailment

possibility is

to evaluate

strength

at

the roof level (i.e. with

force) with nominal minimum reinforcement ratio, and curtail the flexural reinforcement

linearly from the region

of constant elastic

strength towards

the roof level until inter-

secting the level from which figure 6.19 together

[PBM90]
which
was

the nominal minimum

reinforcing is used.
the
same

The demand from the nonlinear time

history analysis of

wa is shown in

with the recommended demand from the shown in

capacity design method

figure

6.12. It is

seen

that above the

plastic hinge

zone,

the demand from the nonlinear time

history analysis exceeds the demand recommended in

occur.

[PBM90]

at two

locations, where plastifications may

The numerical results

confirmed this For die

by indicating limited but clear yielding.

flexible wall with die aspect ratio ra
=

more

8.0 and die fundamental

eigen-

frequency

fi

0.40 Hz, the outcome of the

design approach suggested here

as

is shown in

figure 6.20.
The

The

plastic hinge length Lp was taken

since it is

5.3

(from entire height H divided

=

by six) which is governing

6.21. If the earer

larger than the wall length Ly,

in

4.0

m.

necessity ofthe suggested flexural strength distribution becomes obvious in figure

approach suggested
zone

[PBM90] with

linear strength decrease

one or

direcdy above hinges

would

the

plastic hinge

would have been used here,

over

several

plastic

certainly have developed

the mid and upper storeys, which would The demand from the

violate the fundamental ideas of the nonnear time

capacity design method.

in

history analysis is compared to the recommended demand according to die

figure 6.21.
In this

existing capacity design method [PBM90]

represent
moment

figure,

the shaded

areas are

demand which is not covered if the recommendations of

areas

[PBM90]
or

followed. In these

stretching

over no

less than four storeys,

more

less

uncontroUed plastification wl take

Two basic difficulties
are

place, which could be verified by numerical analyses.

approach of [PBM90].
The

thus identified with the earer

first is the fact that the flexural

overstrength

may

develop

over

large parts

of the

plastic

hinge
This

and may thus transfer

a moment

of that size to the

region intended to remain elastic.

means

that it is necessary

to

provide

for

an

increased

strength at the beginning of the

172

CHAPTER SIX

Design strength with nominal

minimum reinforcement and
zero

normal force

Demand from nonlinear

time

history analysis

Proposed design strength in elastic region

Moment from

equivalent static

force calculation ME

Overstrength Mt Design strength MR >*fRME

(MNm)

30 25 20 15 10 5

0
Hz

^=0.67

Figure

6.18

Suggested strength distribution for eight-storey wall,

aspect ratio ra

5.33

Design strength
minimum
zero

with nominal'

reinforcement and
Demand from

normal force

capacity design rules according to [PBM90], [PP92]

Demand from nonlinear time

history analysis taking into account steel strain hardening

Moment from equivalent static force calculation ME

6.0 m

Design strength MR
^M*^
25 20 15 10 5

>

yR ME

(MNm)

f1=

0.67 Hz

Figure

6.19

Comparison of moment

demand from nonlinear time

history analysis
ra
=

to

demand recommended in the

capacity design method [PBM90J, aspect ratio

5.33

CAPACITY DESIGN CONSIDERATIONS

173

Design strength
zero

with nominal

minimum reinforcement and normal force

Demand from nonlinear

time

history analysis

Proposed design strength in elastic region

Overstrength M

Design strength MR
20 15 10 5

>

yR ME

(MNm)

fi =0.40 Hz

Figure 6.20 Suggested strength distribution for eight-storey wall, aspect ratio ra-8.0

Design strength
minimum
zero

with nominal

reinforcement and

normal force

EE3

Unintended
caused

yielding by moment

demand not covered

Demand from

by existing capacity design rules

'

capacity design rules according to [PBM90], [PP92]

Demand from nonlinear time

history analysis taking into account steel strain hardening

Moment from

equivalent static

force calculation

ME

5.3

Design strength MR >yR ME

m

(MNm)

20 15 10 5
1=

0.40 Hz

Figure

6.21

Comparison of moment demand from

nonlinear time

history analysis
ra
=

to

demand recommended in the

capacity design method [PBM90], aspect ratio

8.0

174

CHAPTER SIX

elastic

region.
to

The increase the

must

be such that the

yield

moment

of that
is
a

cross

section

corresponds
which

overstrength

moment

of the

plastic hinge. This

phenomenon

essentiaUy

has to be

separated

from the behaviour of Upper storeys, and it is not

accounted for in the earer The

next

approach.
shown in
the
a

difficulty pertains to the tacit assumption that the moment demand wl foow
one

a curve

similar to the

figure 6.12.

It is mentioned that

higher modes
wiU of the

may be

alter this curve, but

apparentiy
line with

assumption is that the influence of these following the design

never

larger than

straight

linear decrease

curve

same

figure.
not

As

we

have seen, the mueh that

a

higher modes

may influence the total moment

curve

for

flexible walls

so

hnear decrease of strength

direcy above die plastic hinge is

what the flexural
as

enough.
on

The aspect ratio of die wall does not enter into the discussion

demand distribution wl look like. This

pertains to

the Swiss code

[SIA160]

well

as to

the

capacity design method [PBM90].

Due to this fact it may be worth in

reflecting over the

structural New Zealand;

fact that the walls

as

expected demand as
are

seen

figure

6.12 is
severe

representative for typical

seismicity such
6.12 will
as

they

designed
for

in countries with
a

considerably stiffer,
would be

and for

given

number of storeys, with smaller aspect ratios than

typical

Europe.

The demand in
as

figure

more

typically reflect

dominant first mode behaviour

expected from stiff walls.

be

The conclusion would be that but may

the demand
not

suggested in figure 6.12 may

adequate for New Zealand practice,

generaUy be projected unchanged to typical European practice.

The here

impact of the above results on the design of the flexural reinforcement is discussed
the wa with aspect ratio of 5.33. The flexural

briefly for

strength

in the

plastic
of

hinge region
amounts to

was

computed

at a

strain

at

the

compressive edge equal

the

to

0.0035 and

19.4 MNm. This value

of fy
the
=

was
no

computed using

ordinary assumption
the

design strength
plastic hinge as
Re

460 MPa and

strain

hardening. Using

overstrength factor
above the

Xo we may obtain

proposed necessary flexural strength for the elastic region

XoRp

1.2(19.4)

23.3 MNm

According to the above it is required that the wa in the elastic region reaches a flexural
strength of 23.3
the confined
MNm. This
can

be achieved

by adding a number of flexural bars D20

at

at

zones at

the ends.

Adding
figure

four D20

each end

as

shown in
cross

figure

6.23

gives

an

elastic

strength of about 22.5 MNm. By comparison,

is shown in 6.22. The 22.5 MNm

the

section of the
to

plastic hinge
guarantee that

zone

are not

quite enough

no

yielding wl occur since

the

overstrength which may be transferred was

CAPACITY DESIGN CONSIDERATIONS

ns

6D20

D10/200

Sym

500
-r

1^/2

3000

Figure

6.22 Wall

cross

section

ofplastic hinge

zone

10D20

A+wm
500

D10/200

Sym

Jsection

Lw/2

3000

Figure 6.23

Wall

cross

of constant elastic region above plastic hinge zone

calculated to be 23.3 MNm. However, this difference of around three percent may be tolerable. The
cross

section

strengdiened

with four additional bars D20 at each end wl in this

example

be necessary to

safeguard against uncontroUed development of plastic hinges in proposed that the additional flexural reinforcement bars be practicaUy allow for enough
and the bond. in order to

the mid and upper storeys. It is

arranged

in

U-shaped loop
region

Figure

24

shows the transition

in elevation

between the

plastic hinge

strengdiened elastic region

making

(left) and the vertical section A-A (right) which is indicated in plan in figure upwards
on

23. Each added bar may be continued the

the other side of the U-tum,

up

opposite bar,
on

or

may be

spced
place

at

the other side of the tum,

seen

using

the

sptice length

depending
24

bar diameter prescribed in the code. As

in the vertical section of figure

(right)

it is

important

to

horizontal bars (no less than bar D10) at the inner

corners

ofthe

U-shape. by
the arrangement of the reinforcement

It should noted that

proposed here,
zone

better

opportunity compared
to

to

determine the
more

geometrical

extension of the
as

plastic hinge
a

is

given
of the

previous

vague
now

assumptions such

e.g.

sixth of the total

to

height of

the wall. The extension may

e.g. be chosen to be

exacy equal

the

height

first storey, which may be

advantageous

of a construction

point of view.

176

CHAPTER SIX

Sym
A*---

A-A

flffi
IIE Eiill

Two horizontal

Hill
EllSC

bars D10

U-shaped
barD20

fc
elastic

500

1^/2

3000

300

Figure 6.24 Proposed transition region between

the

plastic hinge and the strengthened

region: elevation (left) and vertical section A-A (right)

It is difficult to suggest

good

rale of thumb for the increased flexural

strength
we

to

be

used

generaUy, based bending

on

die

relatively few computations

the upper end of the

of this report Since

observe

that the

moment at

plastic hinge generaUy

appears to be

sghy

smaUer than at die base, the

proposed
we

value of 20 % is

probably somewhat on the

conservative side. However, since ratios in this respeet,

a

do not know the

impact of e.g. different aspect

value of around 20 % still appears justifiable at this time.

a

Findings

on

the need for

slighy increased flexural strength right above the plastic [Haas93] in his smdy
a

hinge

were

also made

by

Haas

on a

four-storey capacity designed

of about 2.0 Hz,

wall with

an

aspect ratio of 2.7 and

fundamental

frequency

using

among others the model

developed in chapter three

of this report

It is clear that

although the suggested design approach worked

it is not shown how this
an

weU for the two other

to

examples shown here, examples.

approach

would work

generaUy for

It would be desirable with

extensive

parametric study partly

in order

obtain die effective moment distribution and

partly in order to calibrate design parameters.

CAPACITY DESIGN CONSIDERATIONS

177

6.6 Shear Behaviour

The

dynamic

shear forces

developed

at

the base of

multi-storey

structural wall

during

severe

seismic action

display the foUowing characteristics:

sectional shear force

1) The sign of the

cross

changes

more

rapidly

than die

sign of die

bending moment.
2) The magnitude of the
the shear force obtained

cross

sectional shear forces may be

considerably larger than

only from the equivalent static force method. experimental studies, [ES 89],
to

Both these

phenomena

have been verified in several

[SE92]. The physical explanation

second one, may be attributed

for the first characteristic which also has

to

do with the

the several
over

higher

modes which

are

contributing
a

simultaneously
force

to

the deformations

the

height

of the wall, which result in

shear

brought down to the base which changes its sign more rapidly.
to

This is due

the fact that the few lowest modes have the bulk of their inertia forces

concentrated in die upper storeys and

wall. The moments caused when

thereby relatively long lever arms

to

the base of the modes

by

these modes wiU take

periods belonging

to these

plotted versus time. However, the higher modes

giving relatively

have the bulk of their inertia forces

located at mid and lower storeys

smaU moment contributions at the base

but considerable shear contributions.

The effect may be illustrated

by

the

eight-storey

wa used in

chapter

five and

previously in this chapter by transforming the overturning moment at the wa base from
the time

history analysis (figure 5.20)

ductility design.

into the

frequency

domain.

Figure 6.25
(visible

shows

Fourier spectrum of the

overturning moment at
It is
seen

the wall base, for die SIA

that the first mode

ground motion
at

input,

and restricted

around 0.5

Hz obtained from the entire 12 s, i.e.

including
(at
4

die

yielding phase)
same

dominates.

Figure
a

6.26 shows

Fourier spectrum of the base shear of the

two

wall, displaying

considerable contribution from mode

Hz), mode 3 (around 9 Hz), and

even

something

from mode 4

(around 15 Hz).

The second characteristic is base of the wa, and

partly

due to die

developed

flexural

overstrength

at

the

partly to the effect of low centre of gravity of the horizontal inertia just discussed, giving
lower than
a

forces of

higher

modes

resulting

centre

of inertia forces which

may be located

considerably

anticipated by

the inverted

triangulr

force

distribution of the

equivalent static force method.

178

CHAPTER SIX

5
ri

-"

10 f

15

20

[Hz]
moment at
=

Figure design,

6.25 Fourier spectrum SIA

of overturning

wall base, restricted

ductility

ground motion input, aspect ratio ra

533

0.3

0.2

^^10 f

n-^*V-

15

20

[Hz]
restricted

Figure 6.26 Fourier spectrum of base shear,

motion input, aspect ratio ra
=

ductility design, SIA ground

5.33

The focus will here be directed

on

the shear behaviour of the

tests
on

eight-storey
some

wall

presented in chapter five,

for the

and
on

already used in the

flexural behaviour

in that

chapter and to

degree

previous discussion

this chapter.

CAPACITY DESIGN CONSIDERATIONS

179

The base shear obtained sectional shear force of die

was

during

the nonlinear time

history analysis

as

the

cross

macro

element in the plastic

hinge placed dosest to the base,

as

displayed

versus

time in

figures 5.37 and 5.38.

For the restricted

well

as

for the

fll

ductity design,

the maximum base shear

oceurs at

around 3.8 seconds, and the

magnitude

amounts to

about 2.3 MN and 2.1 MN for the restricted and fll

ductility

designs, respectively.
We wiU discuss here the numerical result in relation to the

design

criteria
are

set

for shear

force

by the capacity design

method. The

higher dynamic
to

shear forces

recognised in

the formula for design shear in

[PBM90] according

the

foUowing
(6.13)

K,
where

<M\,>oy
is the

V^,

design

shear force, fi) is the

dynamic ampfication factor, <&0 equivalent

is the

flexural overstrength factor, and

calculation. The formula

VE

is die shear force from the

static force

(6.13) originally arises from recommendations in the New

see

Zealand Standard NZS 3101, The flexural

[NZS3101].
treated earer in this

overstrength

was

chapter,

and for the

dynamic

ampfication factor, [NZS3101] suggests the foUowing expressions

Buildings with up until

six storeys:

cov

0.9 +

(6.14)

Buding

with

more

than six storeys:

<

1.3

<

1.8

(6.15)

In

our

example, formula (6.15) would apply with

ductity design,
the shear force

8, giving cov

1.57.

Starting

with the restricted

VE

from the inverted

was

triangulr force
obtained
as

distribution of the
MN. The flexural

equivalent static force calculation in our example

overstrength developing in
17 percent, walls
was

0.74

in section 6.2 determined to be in

the rnge of 15

to

depending on chosen displacement ductility.

An attempt to confirm the

quantities of the formula (6.13)

and
on

shows that side

we

we

know the left

hand side from the time have estimates of the

manner, the

history analysis,

the

right hand
to

know

VE

and

we

overstrength

reached

according

the numerical model. In this

dynamic magnification
may establish
a

factor may be determined and

compared
as

to

the

proposals (6.14) and (6.15).

At first
we

total shear

magnification

factor ms, defined

the

produet

of the flexural

overstrength factor and the dynamic ampfication factor

180

CHAPTER SIX

a).O0fa,
example
we

(6.16)
obtain

For our

m.

equal

to

2.310.1 A
in

3.11. Values in this rnge may be

found in several of the

examples presented
the time

[SE92]. Inspection of the aetually reached

from

overstrength
5.19) gives
respeet
to

moment
a

during

history analysis (around 22.5 MNm,

sectional flexural
=

figure

numerically obtained strength

the
as

cross

overstrength
the

factor with

available

22.5/19.4

1.16, showing good agreement with the

we

estimations made in this factor with respeet

to

chapter. Here, however,

equivalent

would rather

use

overstrength
=

static moment of 16.1 Nm, which

gives <boa

22.5/16.1

1.40.
we

Using expression (6.16)

This value is

may

now

solve for (o and obtain cov

a

3.11/1.40

2.22.

clearly higher

than the limiter 1.8 for

large

number of storeys

given by

expression (6.15),
of the
same

and in die first

place it is mueh higher than

the obtained value of 1.57

expression using the correct number of storeys equal to eight. The difference
=

is ofthe

magnitude 2.22/1.57

1.41 i.e. 41 percent.

Repeating the procedure for the fu ductity design we obtain VE

gives
m.

as

0.43 MN, which

equal

to

2.1/0.43

4.88. From

figure 5.20
19.0/9.7

the maximum

bending

moment

reached for fll

ductility is about 19.0 MNm, which gives the flexural overstrength with
moment as
=

respeet

to die
=

equivalent static
2.49. By

1.96. Thus (Ov may be solved

as

4.88/1.96
2.49/1.57

comparison

with the

suggested magnification

this value is

1.58, i.e. 58 percent higher.

The actual shear

capacity

of around 2.7 MN for both

designs is only applied

in the

due to minimum

reinforcing requirements
smaller

of the code
not

(which

are

also

capacity design
section had
a

method), and would for example

area.

have been reached if the wall

cross

One could aetually conclude that in die present

examples,

the minimum

reinforcement

requirement

is

more

conservative than the absolute shear

capacity

requirement of expression (6.13). Although only

forces obtained
two

numerical

examples
several

of the shear behaviour

were

presented here,
that the shear

the results indicate,

at

together with

experimental smdies carried out,

the waU base

during nonlinear dynamic

behaviour reaches

large values

which exceed even those

anticipated using conservative design formulae.

the

Keintzel

[Kein88a], [Kein88b] performed elaborate numerical smdies

on

dynamic

shear force demand for section 2.2.1 in this

the

multi-storey walls, using

to

beam element model (as described in

diese smdies, it
was

report)

simulate the wall.

to

During

found that

higher modes contributed considerably

the increased shear forces.

CAPACITY DESIGN CONSIDERATIONS

181

In contrast to the flexural hand calculation

design,

it will

not

be

attempted

here

to

derive

an

improved
model
what

capacity design procedure

of various walls

for shear. Further results from nonlinear time

as

history analyses
presented

using

numerical waU model such

the

macro

in this report would be needed first.

Merely a few

words will be said

as to

might be the result concerning the

hand calculation formula

(6.12) for shear demand upon

completion of such additional analyses.

One

possibility would be
die

to

further increase the value cov for

and

dynamic magnification
give die
and

by modifying

expressions (6.14)

(6.15).

A further refined method would

dynamic ampfication factor directiy

in relation to the

in relation to the natural

frequencies of the wa,

ground motion characteristics.

that

Pardy, the risk

expression (6.13)

is

on

the unconservative side is reduced

by

die

fact that the included flexural

overstrength

is somewhat overestimated, and

partially
for

makes up for die underestimation of the

to

dynamic magnification

factor. The latter appears

in the

be

suggested with

the

high flexural overstrength factor included

in
a

expression
chapter,

the total shear,

resulting overall
the total

fairly realistic
as

estimate ofthe

dynamic

shear forces.

However,

as

indicated in this section

well

as

in the first section of this

the

constituents

forming

dynamic

shear force

display

different relation to each

other, and this should be the subject of further research and considered in future design
formulae.

182

183

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

Summary
The

objective of the study presented in this report is the development of numerical

models

for the Simulation of the behaviour of multi-storey reinforced concrete structural walls in

budings subjected
walls. After
an

to

earthquake action,
a

with

particular emphasis on capacity designed

previous
work in this field, two models

introduction and

review of

working

in

considerably different manner were selected for further development.

macro

The first of these, die so-caed behaviour of

a

model, deals direcy with the

cross

sectional

structural wa

by

means

of nonnear

Springs.
on

An efficient and transparent

version of this type of model is derived, with

conditions. Global

emphasis

fulfling simple
on a

kinematic

hysteretic

rules

are

derived

largely based

knowledge

of the basic

physical

behaviour

complemented

with

empirical

observations. A closed mathematical

form of the stiffness matrix of a

macro as

element is derived.
on

The second model, referred to and

on

the micro model, is based

die mechanics of sods

nonnear material models. The

development of this type of model was carried out

by accommodating
the
macro

the most essential features of reinforced concrete behaviour. As for

model, the development utilised basic physical behaviour and empirical

observations. Different contributions to the

derived in
a

composite

material modulus matrix

were

clear

manner.

Both these models

element code. The material

were

programmed

and

implemented

in

an

existing general finite

simple use of the two

the

models is facitated here

are

by the user element and user

Option.

The models

developed Appendix
a

user-friendly, with input of data

according to speeifications in
Both models had to

of this report

undergo

series of tests, which served the purpose of static tests and gave
an

checking

the

reabity against experimental data from

no

estimate of parameter
behaviour of

influence. Since

suitable

experimental data was avaable for the dynamic

multi-storey was, the micro model had to serve as comparison for the macro model.
After
a

chapter devoted
a

to

reliabity

tests

and tests
some

on

the

Performance

of

capacity
capacity analysed

designed building, design method in

An

further

chapter

deals with

specific problems

of the

view ofthe numerical models

developed here. by
macro

eight-storey capacity designed

attention is

wall is modelled
to

elements and

dynamically. Special

paid

the

ynamic

curvamre

ductity

demand in the

184

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

plastic hinge,

die

dynamic bending strength

moment

and shear force demand. An

improved

distribution of the flexural

over

the

height of the wa is proposed.

Conclusions

In the introduction of this report it

was

concluded that it is difficult to find

numerical

model for
For the
macro

structural wa which is ideal for a of

analysis tasks.
walls of
to

Performance check
was

multi-storey

capacity designed budings,

its

the

model

found

especiay useful.

This is due

realistic

hysteretic behaviour,
of monitoring
cross

capabity

sectional quantities,

especially such as section forces

and curvature
-

ductity, and

mited numerical effort

The

reliabity

tests of the macro model were

performed using experimental

was

data from

static tests. It

was

concluded that die

macro

model

capable

of

simulating

the most

important aspects
have

of the static tests without any

major parameter adjustments.

problems
it

However, for dynamic behaviour no suitable experimental data exists, and in order to
a

comparison

basis for the

macro

model for such

was

concluded that

micro model should be

cases

developed,

which should also

serve as a

complimentary model for

when detaed

analysis is needed and for irregulr geometries.

developed
in this report is first of a

The

macro

model

relatively simple
number of

to compre-

hend. Extensive

knowledge
use.

of the mechanics of solids is not necessary for its under

It consists of

standing
for
a

or

for its basic

only the necessary

The model

spring members

fuU

description of the basic kinematic cases.

or

proved to be an efficient tool

especiay for the case of cyclic

its

dynamic

behaviour. This is attributed to the direct way behaviour. Furthermore, the effect of
can

relatively few components influence the global

on

parameter modifications
model.
The model,
as

the

global behaviour

be

easily followed

in the

macro

implemented and
extensive

described in the

Appendix, is also user-friendly.

degrees
of freedom may be Whe pro

Compared viding

to

the micro model its mueh lower number of

important during
a

dynamic computations involving large models.

the
macro

reasonable
on

global behaviour, however,

damage,
such
as

model does not

provide mueh
macro

Information

locased

crack directions and

yielding.

The

model

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

185

also

inherentiy

assumes a

regulr
wa.

wall geometry,

Symmetrie

cross

section, and

cannot

easy handle openings in

For the micro model obtain
a

developed

in this report it is concluded that it is be

possible

to

clear and

simple

material model for

to

uniformly

reinforced structural walls. It

was

found that it is
concrete

usuay adequate

consider the most

was

important phenomena of reinforced

further shown that it is

in order to obtain reasonable results. It

possible

to

divide the contributions to the material modulus matrix into different parts. The interaction

effects between concrete and reinforcement may be modelled in

which may be its

transparent

manner,

generalised into different levels. The major advantage of the micro model is
of local
an

capabity to give relatively detailed Information

yielding easy
to

damage,

such

as

direction of
in
a

cracks and

of reinforcement. Furthermore,
more

irregularity

or

opening
are

wa

may be

modelled. Some of the model's

cross

important drawbacks
as

its

general

inabity

monitor

sectional

quantities,
demands

such

section forces and curvature

to

ductity,

and its

high computational
black box, the

compared

the

macro

model, which is
use

especiay important during Solution of dynamic problems. Lastiy,

micro model
as a user

in order not to

the

ought to

be famiar with the nonnear behaviour of

materials.

The

Performance

tests on the

example

of a

capacity designed buding indicated that a dynamic analysis, using

some

reasonable structural behaviour may be achieved for nonnear

the

capacity design recommendations

in the current version. However, in

respects

the time

history analysis
curvamre

indicated that

particular care

in the
as a

design

must

be taken. The

dynamic

ductity

demand in the

plastic hinge

function of the wa aspect

in the

ratio and the

displacement ductility,

differs from the

existing suggestions

capacity

design method.
The distribution of flexural

strength

is

one

problematic
by

area, in which the current

design recommendations do
true

not

always provide a conservative Solution. This is especiay

was

in the

case

of slender walls. It

shown

an

energy

study

that

slender wa the
case

dissipates more
for
a

energy in the upper storeys

compared to the plastic hinge than is

on
a

stiffer waU. The

briefly presented proposal

how to

improve

the distribution of but should be

flexural reinforcement tested

more

appeared

to

work well for

two

chosen

examples,

generaUy. history analyses

also

The maximum base shear obtained from the time exceeded the ations.

considerably
recommend

anticipated dynamic

shear force

given

in the

capacity design

186

SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

Recommendations for Future Research

The models

developed
regulr

and tested in this report, have

only

been

presented
an

in their basic

forms, which enabled simple comparisons with experimental data and

behaviour of
wider
a

estimation of the
seen

cantilever wa. The purpose of the numerical models,

in

perspective,

to serve as a tool for structural

design

and for the

analysis of existing

large structures,
models be

could not be dealt with within the scope of this report

Therefore, it is recommended that modifications and further development of the

performed

so as to

enable die

modeUing

of

more

compcated

was found in

real structures, with features such as;

connections

to

frames,

coupling beams, (coupled was)

and was

buding up three dimensional

flexural

cores, such

as

stairways and litt shafts

Regarding the

strength distribution of capacity designed was,

findings

an

extensive

parametric study on improved design parameters should be carried out Attempts should
be made to try to confirm the
on

the effective moment distribution

by further

experimental
The

and numerical studies. Nonlinear time

on

history analyses,
time

in

particular, should

be carried out

designs with improved flexural strength distribution.

of base shear obtained

larger magnification
by
an

during

history analysis should be

confirmed

additional

parametric study
be extracted
as

and

by simple experiments which easy

to

aUow the shear of the wa frame-waU tests.

to

opposed

frequentiy performed complex

187

ZUSAMMENFASSUNG,

SCHLUSSFOLGERUNGEN

UND

AUSBLICK

Zusammenfassung
Das Ziel der

voregenden Arbeit ist die Entwicklung numerischer ModeUe fr die Simula

tion des Verhaltens

mehrstckiger Stahlbetontragwhde Einwirkung

von

in Gebuden, insbesondere

kapa

zittsbemessener Tragwnde unter der und einer Uebersicht ber

Erdbeben. Nach der

Einfhrung
grund-

bisherige
zur

Arbeiten auf diesem Gebiet werden zwei

stzch verschiedene ModeUe

Das erste der beiden

Weiterentwicklung ausgewhlt.

ModeUe, das sogenannte MakromodeU, befasst sich mittels nicht

linearer Federn direkt mit dem

Querschnittsverhalten der Tragwand. Es wurde eine effi

ziente und verstndche Version dieses

ModeUtyps entwickelt

wobei der

ErfUung

ein

facher kinematischer Bedingungen besondere

Beachtung geschenkt wurde. Globale Hys

physikaschen Verhalten,
Fr die wurden jedoch

tereseregeln basieren
auch durch
elementes wurde ein

im wesentlichen auf dem

empirische Beobachtungen erweitert

Steifigkeitsmatrix eines Makro

geschlossener Ausdruck hergeleitet.

Dieser
von

Das zweite ModeU, MikromodeU genannt basiert auf der Kontinuumsmechanik und

auf nichtnearen

Stoffgesetzmodellen.

Modelltyp

wurde unter

Bercksichtigung

der wesentlichsten Effekte des Verhaltens modeU wurden der

Stahlbeton entwickelt Wie beim Makro

Entwicklung sowohl das physikalische Verhalten als auch empirische

Die verschiedenen Anteile der Materialmodulmatrix fr

Beobachmngen zugrande gelegt

Die beiden Modelle wurden

den Verbundwerkstoff Stahlbeton wurden in verstndlicher Weise

hergeleitet
Finite Elemente

programmiert

und in einem

gegebenen

Programm implementiert.
Anhngen A und durchgefhrt,
Da keine
B ist

Die Bentzerfreundlichkeit der beiden ModeUe wird durch die

User Element und die User Material

es

Optionen sichergesteUt Mit Hufe der Angaben in den

mgch, die beiden ModeUe problemlos anzuwenden. Zuverlssigkeit

der ModeUe

Sowohl mit dem Makro- als auch mit dem MikromodeU wurden numerische Testserien
die dazu dienten, die
zu

aufgrund

von

Daten

statischer Versuche

berprfen sowie

die Einflsse der

Modellparameter abzuschtzen.

geeigneten
zur

Versuchsdaten

von

dynamischen Beanspruchungen mehrstckiger

Tragwnde

Verfgung standen,
Kapitel,

dienten die

Ergebnisse dynamischer Berechnumgen

der

mit dem MikromodeU als

Vergleichsbasis
zur

fr das MakromodeU.

Nach einem

das den Tests

Ueberprfung

Zuverlssigkeit numerischer

ModeUe sowie den Tests

bezglich

des Verhaltens eines

kapazittsbemessene Gebudes

188

ZUSAMMENFASSUNG, SCHLUSSFOLGERUNGEN UND AUSBLICK

gewidmet ist,

wird in einem weiteren

Kapitel

auf

einige spezieUe Probleme

der Be
ent

messung mit der

Kapazittsmethode eingegangen.

Diese werden anhand der hier

wickelten numerischen Modelle erlutert

Eine und
im
zu

achtstckige kapazittsbemessene Tragwand wird mit Makroelementen modelert

berechnet Die Wand wird insbesondere auf

dynamisch

Krmmungsduktittsbedarf
Eni

plastischen Gelenk, auf Biegemomenteiner verbesserten

und

Querkraftbedarf untersucht.

Vorschlag

Verteung des Biegewiderstandes ber die Hhe der Wand wird

vorgesteUt

Schlussfolgerungen
In der

Einfhrung

wurde bereits darauf

hingewiesen,

dass

es

schwierig ist,

ein fr alle

Rechenaufgaben ideales ModeU fr Tragwnde zu bden.

Zur

Ueberprfung

des Verhaltens

kapazittsbemessener Tragwnde
und
zwar

wird das Makro

modeU als besonders

-

geeignet erachtet

aufgrund

seines realistisches seiner Fhigkeit

Hystereseverhaltens und

Querschnittsgrssen, insbesondere Schnittgrssen und

Krmmungsduktitt wiederzugeben, sowie

-

wegen seines

begrenzten Rechenaufwandes.
wurden

Die

Zuverlssigkeitsberprfungen des Makromodelles

durchgefhrt.
Es wurde

aufgrund

von

Ver

suchsdaten statischer Versuche

festgesteUt dass

das Makro

modeU

fhig ist

die

wichtigsten Aspekte
zu

der statischen Versuche ohne wesendiche Para-

metermodifikationen
We
vom

simulieren.
Verhalten

dynamischen

mehrstckiger Tragwnde

kerne

geeigneten Ver

suchsdaten voregen, wurde ein MikromodeU entwickelt das als Vergleichsbasis bei der

Beurteung dynamischer Berechnungen

fr detaerte Das MakromodeU, dessen

dienen soU, und das auch als zustzliches ModeU verwendet werden kann.

Berechnungen und unregelmssige Geometrie Entwicklung

hier beschrieben wird, ist in erster Linie gut

verstndlich. Vertiefte Kentnisse der Mechanik kontinuierlicher Medien sind weder fr das Verstehen des Modes noch fr dessen

Anwendung notwendig.

Das MakromodeU
Beschrei
er

besteht

aus

der absolut

notwendigen Anzahl Elemente, die fr eine umfassende Werkzeug, insbesondere

im Fall

bung

der

grundlegenden kinematischen Bedingungen notwendig sind.

zyklischen

Das ModeU

weist sich als effizientes

und

dynamischen

Verhaltens. Der Grand dafr ist dass die

vergleichsweise wenigen Elemente das globale

von

Verhalten direkt beeinflussen. Im weiteren ist der Einfluss auf das

Parametermodifikationen

globale Verhalten meist gut vorhersehbar.

ZUSAMMENFASSUNG, SCHLUSSFOLGERUNGEN UND AUSLBUCK

189

Das MakromodeU, wie

es

implementiert

und im

Anhang

A beschrieben

wurde, ist

weiter besonders bentzerfreundch.

Bezogen auf das MikromodeU ist die normalerweise

bei umfassenden

betrchtlich kleinere Anzahl

schen

Freiheitsgrade ausschlaggebend Modelerung

grosser

dynami

Rechenaufgaben,

die sich bei der

Systeme
es

steen. Whrend
nicht in der

das MakromodeU ein reastisches

globales

Verhalten

zeigt

ist

dagegen

Lage,

genaue Informationen ber lokalisierte Schden wie

Rissrichtungen

und lokales

Fessen der

Bewehrung

zu

ermitteln. Das MakromodeU baut im weiteren auf Vorraus

setzungen wie gleichmssige Wandgeometrie und symmetrischer

und ist

Wandquerschnitt
Oeffnungen,

auf
zu

demzufolge

auch nicht in der

Lage, Abweichungen,

wie z.B.

erfassen. Das MikromodeU rialmodeU fr dass

es

zeigt

dass

es

auch

mgch ist
die

ein relativ klares und einfaches Mateentwickeln. Es stete sich heraus,

Verhaltens des
zu

gleichmssig

bewehrte

Tragwnde

zu

in den meisten FUen

gengt

nur

wichtigsten Phenomne des

um

Stahlbetons in die Betrachtung einzubeziehen,

erhalten. Es wurde weiter

realistische

globale Ergebnisse

gezeigt

dass

es

gut mgch ist, die verschiedenen Antee der

Materialmodulmatrix in separater Form darzustellen. Die Interaktionseffekte zwischen Stahl und Beton knnen in einer klaren Weise modelliert und auf verschiedenen Stufen

generalisiert richtungen
knnen

werden. Der
es

Hauptvorte betreffend

die

Ergebnisse Uegt geben kann,

fr das Mikro wie z.B. Risse Weiterhin

modeU darin, dass

viel Information ber lokale Schden

und lokales Fessen der

Bewehrung in verschiedenen Richtungen. egen darin,

geometrische Abweichungen und Oeffnungen mit Leichtigkeit modelert werden.

der

Einige
keine dass

schwerwiegenden

Nachtee des Mikromodees

dass

es

i.a. und

Querschnittsgrssen
meistens

wie Schnittkrfte und

Krmmungsduktilitt wiedergibt,
was

es

verhltnismssig

rechenintensiv ist,

insbesondere bei der

Lsung
wenn

umfangreicher dynamischer Probleme wesentch gewisse grundlegende Kentnisse der nichtnearen

Die

ist Im weiteren

gut auch,

dass

das MikromodeU nicht als eine Black Box bentzt werden soU, und

so muss

der Anwender

orthotropen Mechanik besitzen.

kapazittsbemessener
ein

durchgefhrten Ueberprfungen bezglich

des Verhaltens

Tragwerke deuten
mit den

darauf hin, dass whrend einer nichnearen der Methode der

dynamischen Berechnung gutmtiges

dass hinsind. Der

heutigen Empfehlungen
der

Kapazittsbemessung

Strukturverhalten erreichbar ist.

Die

Ergebnisse

Zeitverlaufsberechnungen

deuten

allerdings darauf hin,

sichich der

Bemessung noch einige weitergehende Ueberlegungen anzusteen

dynamische Krmmungsduktittsbedarf im plastischen Gelenk,

schlankheit und der

als Funktion der Wand

gewhlten Verschiebeduktitt, unterscheidet sich von den bisherigen Kapazittsbemessung.

und die Die bisher bliche

Angaben

der Methode der kann

Verteung

von

Biegekapazitt

problematisch sein,

heutigen Bemessungsempfehlungen

190

ZUSAMMENFASSUNG, SCHLUSSFOLGERUNGEN UND AUSBLICK

ermgchen
FaU bei in
emer

nicht immer eine

Lsung

auf der sicheren Seite. Dies ist insbesondere der

ausgeprgt schlanken Tragwnden. Durch eine Energiestudie wurde gezeigt, dass

flexiblen Wand in den oberen Stockwerken im
als in einer

Vergleich zum plastischen Gelenk

daher steifen Wand. Der kurz

mehr

Energie dissipiert wird

Vorschlag
zur

gedrungenen und
aber noch

beschriebene das

einer verbesserten Verteung der

Biegebewehrung

scheint fr

gezeigte Beispiel gut zu funktionieren, sote

zu

genereller getestet werden.

Im weiteren ist

erwhnen, dass die

aus

der

Zeitverlaufsberechnung resultierende

maximale

Querkraft am

Wandfuss die in den

Empfehlungen der Methode der Kapazitts

bemessung angegebenen Werte deudich bersteigt

Ausblick

Die in diesem Bericht entwickelten Modelle erscheinen die einfache

nur

in einer grundlegenden Form,

Vergleiche

zu

Versuchsergebnissen

und

gewisse Schtzungen des Verhaltens

emer

eines

Tragwandkragarmes ermgchen.
ein

Der Zweck dieser ModeUe, in

erweiterten und fr

Perspektive gesehen,
Arbeit noch nicht

Werkzeug

darzustellen fr

Bemessungsuntersttzung

Berechnungen von gegebenen

Es wird deshalb
weitere

grossen Strukturen, konnte innerhalb des Rahmens dieser

vostndig realisiert werden.

empfohlen, die notwendigen Modifikationen

dieser ModeUe durchzufhren, wie sie in reellen Gebuden
an

vorzunehmen und eine

von

Entwicklung

so

dass das ModeUieren

kompti-

zierteren

Tragwnden,

gefunden werden, knftig realisierbar

wird. Besonders
-

gedacht wird
zu

Tragwnde

mit mit

Koppelung

Rahmensystemen,

Koppelungsriegeln (gekoppelte Tragwnde),

ausgebdet als dreidimensionale Kerne, wie z.B. Treppenhuser und Liftschchte.

Hinsichich der Verteung der

Biegekapazitt von kapazittsbemessenen Tragwnden,

durchgefhrt werden.
durch Es sote versucht werden, und weitere
nume

sote eine umfassende Parameterstudie

die effektive nichtlineare

Momentenverteilung

experimenteUe

rische Studien

zu

besttigen.

Insbesondere soten Wnde mit verbesserter

Biegekapaziresul

tts-Verteung durch nichineare Zeitverlaufsberechnungen analysiert werden.

Die grssere

Amphfikation der Querkrfte, die aus den Zeitverlaufsberechungen durchgefhrt werden,

die

tierte, sote durch eine weitere Parameterstudie berprft werden. Ebenso soten dyna
mische Versuche
messen, was
es

erlauben, die Querkraft

an

der Wand

zu

bei den

hufig vorkommenden komplizierten gemischten Wand-Rahmen

Versuchskrpern nicht mglich ist

191

REFERENCES

[Abaq89] [Abaq91]
[Abra91]

Abaqus, Theory Manual, Hibbitt,

Rhode Island, 1989

Karlsson & Sorensen, Inc., Providence,

Abaqus,

Users' Manual, Hibbitt, Karlsson & Sorensen, Inc., Pawtucket,

Rhode Island, 1991

Abrams, D.P. "Laboratory Definitions of Behaviour for Structural Compo nents and Buding Systems", Earthquake-Resistant Concrete Structures, Inelastic Response and Design, S.K. Gosh, Editor, ACI Special PubU cation SP-127, American Concrete Institute, Detroit 1991, pp. 91-152

[ACISP84]

Earthquake Effects on Reinforced Concrete Stractures, US-Japan Research, J.K. Wight, Editor, ACI Special Publication SP-84, American
Concrete Institute, Detroit, 1984

[Bach67]

Bachmann, H., "Zur plastizittstheoretischen Berechnung statisch unbe stimmter Stahlbetonbalken", Dissertation, Institut fr Baustatik, Eidge nssische Technische Hochschule (ETH), Zrich, 1967

[Bach89]

Bachmann, H.

et

al, "Die Erdbebenbestimmungen der Norm SIA 160",

Dokumentation, D 044, Schweizerischer Ingenieur- und Architekten-

Verein, Zrich, 1989

[Bach91]

Bachmann, H., "Earthquake Design of Bridges, The Swiss Code App roach", European Earthquake Engineering Journal, No. 3., 1991, and Bericht Nr. 187, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule (ETH), Zrich, 1992

[Bach93]

Bachmann, H., "Erbebensicherang von Bauwerken", Vorlesungs-Autographie, Institut fr Baustatik und Konstruktion, Eidgenssische Techni sche Hochschule (ETH), Zrich, 1993 Bertero, V.V., Aktan, A.E.,

[BACS84]

Simulation Tests and Associated Studies of a l/5th Scale Model of a 7Story Reinforced Concrete Test Structure", US-Japan Cooperative Earth quake Research Program, Report No. UCB/EERC-84/05, Earthquake Engineering Research Center, University of California, Berkeley, 1984

Chamey, F.A., Sauce, R., "Earthquake

[Bath82]

Bathe, K.-J., "Finite Element Procedures in Engineering Analysis", Prentice-HaU Inc., Englewood Cliffs, New Jersey, 1982

[BG80]

Bazant, Z.P., Gambarova, P.G., "Rough Cracks in Reinforced Concrete",

J. Struct. Div., ASCE, Vol. 106, No. 4,

April, 1980,

pp. 819-42

[BG84]

Bazant, Z.P., Gambarova, P.G., "Crack Shear in Concrete, Crack Band Micro-Plane Model", J. Struct. Eng., ASCE, Vol. 110, No. 9, Sept, 1984, pp. 2015-35 Bachmann, H., Linde, P "Dynamic Ductity Demand and Capacity Design of Earthquake-Resistant Reinforced Concrete Walls", Proc. Tom Paulay Symposium, University of California, San Diego, Sept. 20-21, 1993, ACI Special PubUcation, American Concrete Institate, Detroit 1993

[BL93]

192

REFERENCES

[BP90]

Bachmann, H., Paulay, T., "Kapazittsbemessung

von

Stahlbetontrag

wnden unter Erdbebeneinwirkung", Beton- und Stahlbetonbau, 85, 1990, H.ll, und Bericht Nr. 177, Institut fr Baustatik und Konstruktion,

Eidgenssische Technische Hochschule (ETH), Zrich, Nov., 1990

[BW91]
Bolander, J., Wight J.K., "Finite Element Modeling of Shear-WallDominant Buildings", J. Struct. Eng., ASCE, Vol. 117, No. 6, June,
1991

[BWL92a]

Bachmann, H., Wenk, T., Linde, P., "Nonlinear Seismic Analysis of

Hybrid Reinforced Concrete Frame WaU Buildings", Nonlinear Seismic Analysis of Reinforced Concrete Buildings, Fajfar, P., Krawinkler, H., Editors, Elsevier Applied Science, London, 1992
[BWL92b]
Bachmann, H., Wenk, T., Linde, P., "Seismic Ductility Demand in

Gravity
[CB89]

Load Columns in Reinforced Concrete Frame-Wall

Budings",

Proc. Wth. WCEE, Madrid, 1992, A.A. Balkema, Rotterdam, 1992

Chappuis, P., Bachmann, H., "Concrete material model for earthquake loading including strain rate and loading history", Bericht Nr. 172, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule
(ETH), Zrich, Juni, 1989
Cervenka, V., "Inelastic Finite Element Analysis of Reinforced Concrete Panels", dissertation, Dept of Civil Engineering, University of Colorado, Boulder, 1970
Cervenka, V., Gerstie, K.H., "Inelastic Analysis of Reinforced Concrete Panels: Theory", Publ. IABSE, Vol. 31-11, Zrich, 1971, pp. 31-45

[Cerv70]

[CG71]

[Chap87]

Chappuis, P., "Modelisation non-lineaire du comportement du beton sous des solcitations dynamiques", Dissertation, Bericht Nr. 155, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule (ETH), Zrich, Juni, 1987
Chen, W.F., "Plasticity in Reinforced Concrete", McGraw-Hl, New York, 1982
Cornelissen, H.A.W., Hordijk, D.A., Reinhardt, H.W., "Experimental
determination of crack-softening characteristics of normal weight and ght weight concrete", Heron, Delft University of Technology, Vol. 31, No. 2, 1986

[Chen82]
[CHR86]

[Colo93]

Colotti, V., "Shear behavior of RC Structural Walls", /. Struct. Eng., ASCE, Vol. 119, No. 3, March, 1993 Cervenka, V., Pukl, R., Eligehausen, R., "Computer Simulation of anchoring technique in reinforced concrete beams", Proc. Int. Conf. Computer Aided Analysis and Design of Concrete Structures, Part I, N. Bicanic et al, Editors, Pineridge Press, Swansea, 1990

[CPE90]

[Curb87]

Curbach, M., "Festigkeitssteigerung von Beton bei hohen Belastungs geschwindigkeiten", Dissertation, Heft 1, Schriftenreihe des Instituts fr
Massivbau und

Baustofftechnologie,

Universitt Karlsruhe, 1987

a

[deBN85]

deBorst, R., Nauta, P., "Non-orthogonal cracks in element model", Eng. Comp., Vol. 2, March, 1985

smeared finite

REFERENCES

193

[Ding85]

Dinges, D., "Vergleichende Untersuchungen von Stahlbetonflchentragwerken unter Bercksichtigung physikalischer und geometrischer Nichtnearitten", Dissertation, Fachbereich Konstruktiver Ingenieurbau, Tech
nische Hochschule Darmstadt, 1985

[DO90]

Dahlblom, O., Ottosen, N.S, "Smeared Crack Analysis using Generazed Fictitious Crack Model", /. Eng. Mech., ASCE, Vol. 16, Jan., 1990

[Drr80]

Drr, K., "Ein Beitrag

besonderer bereich 1980

zur

Berechnung

von

Stahlbetonscheiben

unter

Bercksichtigung des Verbundverhaltens", Dissertation, Fach Konstruktiver Ingenieurbau, Technische Hochschule Darmstadt
to

[ES89]

Eberhard, M.O., Sozen, M.A., "Experiments and Analyses

study the Response Yielding Columns", dissertation, Report UILU-ENG-89-2007, Department of Civ Engineering, University of Dnios, Urbana-Champaign, 1989
of Reinforced Concrete Frame-Wa Stractures with

[FF91]

Fajfar, P., Fischinger, M.,

Struct. 1991

"Mathematical

structural walls for nonnear seismic

modeling of reinforced concrete analysis", Proc. Europ. Conf. on

A.A. Balkema
,

Dynamics, Eurodyn '90, Bochum,

Rotterdam,

[Fran70]

Franklin, H.A., "Nonlinear Analysis of Reinforced Concrete Frames and Panels", dissertation, Report No. UC-SESM 70-5, Dept. of Civ Engi neering, University of California, Berkeley, 1970

[FVSFZD90] Fischinger, M., Vidic, T., Selih, J., Fajfar, P., Zhang, H.Y., Damjanic, F., "Validation of a macroscopic model for cycc response prediction of RC was", Proc. 2nd int conf. on Computer aided analysis and design of concrete structures, Zell am See, April 4-6, 1990, Bicanic, N. Editor, 1990 Swansea Pineridge Press,
,

[Glem84]

Glemberg, R., "Dynamic Analysis of Concrete Stractures, Constitutive ModeUing, Dynamic Substructuring, Computer Implementation", disser tation, Publication 84:1, Department of Structural Mechanics, Chalmers University of Technology, 1984 Gupta, A.K, Maestrini, S.R., "Tension Stiffness Model for Reinforced Concrete Bars", J. Struct. Eng. ASCE, Vol. 116, No. 3, March, 1990
Glemberg, R., Samuelsson, A.,
crete

[GM90]
[GS83]

"A General Constitutive Model for Con

Structures", Nordic Concrete Research, PubUcation 2, The Nordic Concrete Federation, Oslo, 1983

[Gust85]

Gustafsson, P.J. "Fracture Mechanics Smdies of Non-yielding Materials

like Concrete:

Modelling

of Tensile Fracture and

Applied Strength

Analysis", dissertation, Report TVBM 1007, Division of Building Materials, Lund Institute of Technology, 1985

[GyU83]

Gylltoft, K.,
Engineering,

"Fracture mechanics models for

fatigue

in concrete

stractures", dissertation, Report 1983:D, Division of Structural

Lulea

University

of Technology, 1983

[GyU89]

Gylltoft, K., "Cyclic Loading", in Ch. 6: "Other Aspects of Material Modelling", Fracture Mechanics of Concrete Structures, Elfgren, L., Editor, RILEM Report, Chapman and Hau, London, 1989, pp. 172-187

194

REFERENCES

[Haas93]

Haas, M., "Erdbebenberechnung

verschiedenen Verfahren",

von Stahlbeton-Tragwandsystemen nach Nachdiplomarbeit, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule (ETH), Zrich,

1993

[Hass91]

zone response to simultaneous normal and shear displacement", RILEM Proc. 6, Analysis of Concrete Structures by Fracture Mechanics, Elfgren, L., Shah, S.P., Editors, Workshop dedicated to Prof. A. HiUerborg, Abisko, Chapman and Hall, London, 1991

Hassanzadeh, M., "Considerations regarding fracture

[Hass92]

Hassanzadeh, M., "Behaviour of fracture process

enced

zones

by simultaneously apped

normal and shear

in concrete infludisplacements", disser

tation, Report TVBM 1010, Division of Buding Materials, Lund Institute of Technology, 1992

[HHT77]

Hilber, H.M., Hughes, T.J.R., Taylor, R.T., "Improved Numerical

Dissipation Eartq. Eng.

[Hb76]

for Time
& Struct.

Integration Algorithms in Structural Dynamics", Dyn., Vol. 5, pp. 283-292,1977

Hilber, H.M., "Analysis and Design of Numerical Integration Methods in Structural Dynamics", dissertation, Report No. UCB/EERC-76/29, Earth quake Engineering Research Center, University of California, Berkeley,
1976

[Hiro75]

Walls and

Hirosawa, M. "Past Experimental Results on Reinforced Concrete Shear Analysis on Them", Structural Research Pubcations, No. 6, 1975, Building Research Institute, Tsukuba, Ministry of Construction,

Japan, [HMM91]

1975

Hayami, Y., Miyashita, T., Maeda, T., "Nonlinear Analysis of Shear Walls subjected to Cycc In-Plane Shear Loading", internal report, Infor mation Processing Center, Kajima Corporation, 2-7, 1-chome, Motoakasaka, Minato-ku, Tokyo, 1991 Hillerborg, A., Modeer, M., Petersson, P.E., "Analysis for crack formu lation and crack growth in concrete by means of fracture mechanics and
finite elements", Cement and Concrete

[HMP76]

Resesach, 6,1976

[Hugh87]
[Kein88a]

Hughes, T.J.R., "The Finite Element Method, Linear Static and dynamic Finite Element Analysis", Prentice-Hall, Englewood Ctiffs, N.J., 1987
Keintzel, E., "Zur Querkraftbeanspruchung von Stahlbeton-Wandscheiben unter Erdbebenlasten", Beton- und Stahlbetonbau 83, H. 7,1988

[Kein88b]
[K0U88]

Keintzel, E., "Vereinfachte Ermitung der Querkraft aus Erdbebenlasten in Stahlbeton-Wandscheiben", Beton- und Stahlbetonbau 83, H. 12,1988

Kollegger, J., "Ein Materialmodell fr die Berechnung von Stahlbetonflchentragwerken", Dissertation, Fachbereich Bauingenieurwesen, Ge
samthochschule Universitt Kassel, 1988

[Kolm85]

Kolmar, W. "Beschreibung der Kraftbertragung ber Risse in nichtnearen

Finite-Element-Berechnungen von Stahlbetonhochbauten" Disser

tation, Fachbereich Konstruktiver Ingenieurbau, Technische Hochschule Darmstadt, 1985

REFERENCES

195

[KP71]

Struct. Div.

Kent D.C, Park, R., "Flexural Members with Confined Concrete", J. ASCE, Vol 97, No. 7, July, 1971, pp. 1969-90

[KSOA82]

Kabeyasawa, T., Shiohara, H., Otani, S., Aoyama, H., "Analysis

of the

Fu-Scale Seven-Story Reinforced Concrete Test Structure: Test PSD-3", Proc. 3rd Joint Technical Coordinating Committe, US-Japan Cooperative

Earthquake
1982

Research

Program, Building

Research Institute, Tsukuba,

[Kupf69]

Kupfer, H., "Das Verhalten des Betons unter zweiachsiger Beanspruch ung", Bericht Nr. 18, Lehrstuhl Massivbau, Technische Hochschule,
Mnchen, 1969

[Lars59]

"Bearing Capacity of Piain and Reinforced Concrete Walls", dissertation, Doktorsavhandngar vid Chalmers Tekniska Hgskola Nr. 20, Gumperts Frlag, Gteborg, 1959
Larsson. L.E.,

[Lind88]

Linde, P., "Experimental and Analytical Modeng of R/C Shear Was", Engineer's thesis, Department of Civil Engineering, Stanford University,
1988

[Lind89]

Linde, P., "Analytical Modeling Methods for R/C Shear Walls", Report No. TVBK-1005, Department of Structural Engineering, Lund Institute of

Technology, 1989
[MB87]
Moazzami, S., Bertero, V.V., "Three-Dimensional Inelastic Analysis of Reinforced Concrete Frame-WaU Structures", Report No. UCB/EERC87/05, Earthquake Engineering Research Center, University of California,

Berkeley, 1987
[MKHH91]
Meskouris, K., Krtzig, W.B., Hansktter, U., Hoffmeister, P., "Nicht
lineares seismisches Tragverhalten ebener und rumcher Stahlbeton tragwerke", Arbeitsbericht fr die Jahre 1989/1991, Wissenschaftche Mitteilungen, SFB-151-Berichte Nr. 20, Sonderforschungsbereich Trag werksdynamik, Ruhr-Universitt Bochum, 1991

[Moaz91]

Moazzami, S., "An Analytical Model Designed to predict the Behavior of Shear Walls", Proc. Int. Conf. Buildings with Load Bearing Concrete Walls in Seismic Zones, Paris 1991, pp. 359-67
,

[MP90]

Moser, K., Paulay, T., "Kapazittsbemessung von erdbebenbeansprachten Stahlbetonrahmen", Schweizer Ingenieur und Architekt, Nr. 44, Nov. 1990, und Bericht Nr. 180, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule (ETH), Zrich, Nov., 1990 Nson, A.H., "Nonnear Analysis of Reinforced Concrete by the FEM", ACI Journal, Vol. 65, No.9, American Concrete Institute, Detroit, 1968, pp. 757-66
Nooru-Mohamed, M.B., "Mixed-mode fracture of
mental
concrete: an

[Ns68]

[Noor92]

experi

approach", dissertation, Dept of Civ Engineering, sity of Technology, 1992

[NS67]

Delft Univer

Ngo, D., Scordelis, A.C.,

Analysis of Reinforced Vol. 64, No.3, American Concrete Concrete Beams", ACI Journal Institute, Detroit 1967, pp. 152-63
,

"Finite Element

196

REFERENCES

[NZS3101]

NZS 3101: Part 1, Commentary NZS 3101: Part 2: New Zealand Code of Practice for the Design of Concrete Structures, Standard Association of New Zealand,

Wellington,

New Zealand, 1982

[01de85]

Analysis of Tensile Fracturing Stractu res", licentiate thesis, Report 1985:010L, Division of Structural Engi neering, Lulea University of Technology, 1985

Oldenburg, M.,

"Finite Element

[OTTKW89]

Ohomori, N., Tsubota, H., Inoue, N., Kurihara, K., Watanabe, S.,
"Reinforced Concrete Membrane Elements
In-Plane Shear Stress", Nuclear pp. 61-72

Subjected to Reversed Cycc Engineering and Design, Vol. 115,1989,

[PBM90] [PL74]

Paulay, T, Bachmann, H., Moser, K., "Erdbebenbemessung von Stahl betonhochbauten", Birkhauser Verlag, Basel, Boston, Beruh, 1990 Paulay, T., Loeber, P.J., "Shear transfer by aggregate interlock", Shear in Reinforced Concrete, ACI Special Publication SP-42, American Concrete
Institute, Detroit, 1974, pp. 1-15 Park, R., Paulay, T., "Reinforced Concrete Structures" John Wiley & Sons, New York, 1975

[PP75]

[PP92] [Prui85]

Paulay, T., Priesey, M.J.N., "Seismic Design of Reinforced and Masonry Buildings" John Wiley & Sons, New York, 1992

Concrete

Pruijssers, A.F., "Description of the stiffness relation for mixed-mode fracture problems in concrete using the rough-crack model of Walraven", Report No. 5-85-2, Stevin Laboratory, Department of Civ Engineering, Delft University of Technology, 1985 Pruijssers, A.F., "Aggregate Interlock and Dowel Action under Monotonie and Cycc Loading", dissertation, Department of Civ Engineering, Delft University of Technology, 1988
Rashid, Y.R., "Analysis of prestressed concrete pressure vessels", Nuc lear Engineering and Design, Vol. 7, No. 4,1968, pp. 334-44
Reinhardt, H.W., Cornessen, H.A.W., "Post-Peak Cyclic Behaviour of
Concrete in Uniaxial Tense and

[Prui88]

[Rash68]
[RC80]

Compressive Loading", Cement and Concrete Resaerch, 14,1984, pp. 263-270

Reinhardt H.W., "Fracture mechanics of an elastic softening material like concrete", Heron, Delft University of Technology, Vol. 29, No. 2,1984
Rothe, D., Knig, G., "Behavior and Modelng of Reinforced Concrete Structural Wall Elements", Proc. 9th. WCEE, Kyoto-Tokyo, Vol. VI, pp. 47-55, 1988

Alternating

Tensile and

[Rein84]

[RK88]

[Rots84]

Rots, J.G., "Variable reduetiefactor voor schuifweerstand van gesheurd beton", Report No. BI-84-33, Institute TNO for Building Materials and Stractures, Delft, 1984 Rots, J.G., "Computational Modeling of Concrete Fracture", dissertation,

[Rots88]
[RR92]

Dept

of Civ

Engineering,

Delft

University

of Technology, 1988

Rule, W.K., Rowlands, R.E., "Predicting Behavior of CyccaUy Loaded RC Stractures", J. Struct. Eng., ASCE, Vol. 118, No. 2, Febr, 1992

REFERENCES

197

[Saat91]

Saatcioglu, M., "Modeling Hysteretic Force-Deformation Relationships Earthquake-Resistant Concrete and Inelastic Structures, Response Design, S.K. Gosh, Editor, ACI Concrete Institute, Detroit, 1991, American PubUcation SP-127, Special
for Reinforced Concrete Elements",

pp. 153-198

[SB90]

Sedarat, H., Bertero, V.V., "Effects of Torsion on the Linear and Nonlinear Seismic Response of Stractures", Report No. UCB/EERC90/12, Earthquake Engineering Research Center, University of California,

Berkeley,
[SE92]

1990

Sozen, M.A., Eberhard, M., "Design Shear for Earthquake-Resistant Walls", Nonlinear Seismic Analysis and Design of Reinforced Concrete Buildings, Suppl. Proc. of a Workshop held in Bled, Slovenia, July 1316,1992, J. A. Blume Earthquake Engineering Center, Report No. 103,

Dept
[SIA160] [SIA162]

of Civ

Engineering, Stanford University, 1992 Ingenieur-

SIA Norm 160, Einwirkungen auf Tragwerke, Schweizerischer und Architekten-Verein, Zrich, 1989

SIA Norm 162, Betonbauten, Schweizerischer

Ingenieur-

und Archi

tekten-Verein, Zrich,1988

[Skri82]

Skrikerud, P., "ModeUe und Berechnungsverfahren fr das Risseverhalten

von

unarmierten Betonbauten unter Erdbebenbeanspruchung", Disserta tion, Institut fr Baustatik und Konstruktion, Eidgenssische Technische

Hochschule (ETH), Zrich, 1982

[T055]

Tomii, M., Osaki, Y., "Shearing Resistance of Reinforced Concrete Bracing Was", Transactions, Architectural Institute of Japan, No. 51, September, 1955, pp.83-95

[VB87]

Vulcano, A., Bertero, V.V., "Analytical Models for Predicting the Lateral

Response

of R C Shear Walls: Evaluation of their

Reabity", Report No.

UCB/EERC-87/19, Earthquake Engineering Research Center, University of California, Berkeley, 1987

[VBC88]

Vulcano, A., Bertero, V.V., Colotti, V., "Analytical Models of R/C Structural Was", Report No. UCB/EERC-88/17, Earthquake Engineering Research Center, University of California, Berkeley, 1988
Vallenas, J.M., A., Bertero, V.V., and Popov, E.P., "Hysteretic Behavior of Reinforced Concrete Was", Report No. UCB/EERC-79/20, Earthquake Engineering Research Center, University of California, Berkeley, 1979

[VBP79]

[Walr80]

Wahaven, J.C, "Aggregate interlock, a theoretical and experimental analysis", dissertation, Department of Civ Engineering, Delft University
of Technology, 1980

[WB90]

Wenk, T., Bachmann, H. "Ductility demand of 3-D reinforced concrete frames under seismic excitation", Proc. Europ. Conf. on Struct. Dynamics, Eurodyn '90, Bochum, A.A. Balkema Rotterdam, 1991, pp.
,

537-41

[Wenk90]

Wenk, T, "Duktittsanforderangen bei Stahlbetonbauten

einwirkung", DAfStb-Forschungskolloquium,

unter Erdbeben Deutscher Ausschuss fr

198

REFERENCES

22./23. Mrz, 1990

Stahlbeton, Eidgenssische Technische Hochschule (ETH), Zrich,

[Wenk93]

Wenk, T., "Nichineare Berechnung

von Stahlbetonrahmen unter Erd Dissertation in bebeneinwirkung", Vorbereitung, Institut fr Baustatik und Konstruktion, Eidgenssische Technische Hochschule (ETH), Zrich, 1993

[WK85]

Wallace, B.J. Krawinkler, H., "Small-Scale Model Experimentation on R/C Assembes", US-Japan Cooperative Earthquake Research Program, Report No. 74, J.A. Blume Earthquake Engineering Research Center,

Department of Civ Engineering,

[WLB93]

Stanford

University, 1985

Wenk, T., Linde, P., Bachmann, H., "User Elements Developed for the Nonnear Dynamic Analysis of Reinforced Concrete Stractures", Proc. Abaqus Users' Conference, Aachen, June 23-25,1993, Hibbitt Karlsson & Sorencen, lhc, Pawtucket Rhode Island, 1993 Zienkiewicz, O.C, "The Finite Element Method', McGraw-Hill, London,
1977

[Zien77]
[ZP074]

Zienkiewicz, O.C, Phillips, D.V., Owen, D.R.J., "Finite element analysis of some concrete non-nearities, Theory and examples", Concrete structures subjected to triaxial stresses, Seminar, 17.-19. May, 1974, Bergamo, IABSE, Zrich, 1974
Yanes, L. A., "A Computer Program for Nonlinear Analysis of Rein
forced Concrete Stractures 1982

[Yane82]

Engineer's thesis, Department

of Civil

Subjected to Lateral Load Reversais", Engineering, Stanford University,

Abbreviations: ACI:
ASCE: EERC: IABSE: RILEM: SIA: UCB:
WCEE:

American Concrete Institute American Society of Civ Engineers Earthquake Engineering Research Center (University of California) International Association for Bridge and Strctural Engineering Reunion Int des Labs. d'Essais et de Rech, sur les Materiaux et les Constr. Schweizerischer Ingenieur- und Architekten-Verein University of Cafornia, Berkeley World Conference on Earthquake Engineering

199

NOTATION

Greek

Upper Case

O0
&

Hexural

overstrength factor

Rotation Yield rotation

&y

Greek Lower Case

ct ct.

Ratio of outer vertical

spring area and wa

area,

parameter in time integration

Fraction of length of constant elastic

strength to lengi of elastic region

ctcl

Fraction of yield force of outer vertical

spring for which flexural cracks

are

closing

ctcr
ct*mm
ct.

Ratio of cracked to elastic stiffness of outer vertical

spring
cross

Ratio of global flexural stiffness reduetion of cracked

Ratio of cracked to uncracked shear stiffness

section

ay
ay

Ratio of yielded to elastic stiffness of outer vertical

spring
cross

Ratio of global flexural stiffness reduetion of yielded Shear retention factor, factor in time Axial

section

8a 8h 8he

integration

displacement displacement displacement

Horizontal Shear

8S
8SC 8V
8V i <5vr

Shear
Shear

displacement
displacement at shear cracking

Vertical Vertical Vertical

Yield

displacement
displacement at left element edge
displacement at right element edge

8y
e

displacement
at zero stress

Strain

Eo

Concrete tense strain

Tense strain at

Eots

zero stress

in tension

stiffening

model

Esou
50h

Softening compressive Softening compressive

Concrete strain

strain for unconfined concrete at 50 % strain for confined concrete at 50 %

strength

strength

200

NOTATION

cr
/

Concreto

cracking strain edge

Vertical strain at left element Vertical

Ely
en r
s

yield strain at left element edge

Strain normal to crack direction

Vertical strain at right element edge

Steel strain
Yield strain

Ey

Curvature, angle between local and global coordinate system

Yield curvature

h
r
Yr
K

Shear strain, factor in time

Resistance factor
Form factor in shear

integration

A
Aq

Mesh correction factor

Overstrength factor for reinforcement steel

Global

\~A

displacement ductity ductity

ductity

H
f-e
cov

Curvature
Rotational

Dynamic magnification factor

Reinforcement ratio Reinforcement ratio in x-direction Reinforcement ratio in

p Px

Py
a

y-direction

Stress
Axial tress

O"0

of
af
Ps

Fibre

concrete stress
stress

Fibre steel

Confinement ratio
Tense reinforcement ratio

Pt
Ots

Concrete tense stress between cracks (tension

Horizontal reinforcement ratio

stiffening stress)

Pw

pWh

Horizontal reinforcement ratio of web

Latin

Upper Case

Ac

Concrete

area
area

Ag
Af

Cross sectional gross

Fibre concrete
area

NOTATION

201

Af

Fibre steel

area

As
A si

Reinforcing steel area Reinforcing steel area crossing

Shear reinforcement
WaU WaU
cross cross

inclined surface

Av

area
area

Aw

sectional sectional

AWi

area

inclined at 45
area

degrees

Aws
B

Horizontal reinforcement

of web

Strain-displacement matrix
Global

damping matrix,

transformation matrix

Cd
Ct
D

Overstrength reduetion factor

Factor for equivalent shear stress

Material modulus matrix

Dc

Concrete modulus matrix

Modulus matrix in

Dg
Dia
Di

global coordinate system

Interaction modulus matrix

Modulus matrix in local coordinate system

Dj^ Largest aggregate diameter

Ds
E

Steel modulus matrix

Young's

modulus

Ec

Young's modulus for elastic smaU strain concrete

Ei
Ek

Dissipated energy
Kinetic energy Elastic strain energy
Inelastic

Ese ESie

(irrecoverable) strain energy

Es,ts
Esx

Increased

Young's

modulus for reinforcement steel due to tension

x-direction

stiffening

Young's modulus for reinforcement steel in

Total energy

ESy
Et Ets
Ev
F
F

Young's modulus for reinforcement steel in y-direction Young's modulus for concrete between cracks in opening phase
Viscous energy
Force

Global force vector

Fc Fcs
Ff

Cracking force of outer vertical spring Spring force of central vertical spring
Fibre concrete force Fibre steel force Residual force

Ff Fr

202

NOTATION

Fs

Spring

force of outer vertical

across

spring stiffening spring

Fts

Steel force

cracks due to tension

Fy
G

Yielding

force of outer vertical

Modulus of rigidity

Gf
H

Fractore energy

Buding height
WaU

Hw
/

height
section

Moment of inertia Moment of inertia of cracked Moment of inertia of elastic

cross

laIe

cross

section

Iy
K

Moment of inertia about strong axis Global stiffness matrix Stiffness of left vertical outer spring

Ki

K2 K3
Ka Kc Kcs Kt

Stiffness of right vertical outer spring

Stiffness of central vertical

spring
of outer vertical

Axial stiffness of horizontal beam

Elastic

compressive stiffness

spring

Elastic

spring stiffness of central vertical spring spring stiffness

of outer vertical

Form factor for shear cracking

Ks Kse
Ku Kve

Horizontal

Elastic

spring stiffness

spring

Unloading stiffness

of outer vertical

spring

Elastic stiffness of horizontal

(shear) spring

Ky
L

Yield stiffness of outer vertical

spring
area

Buding length, length

Length of elastic region Length
Wa

of integration

Le Lec

of region of constant elastic

strength

Lp
Ly,
M M

Length of plastic hinge zone length

moment

Bending
Global

mass

matrix

ME
Mi

Bending moment from static equivalent force Bending moment demand

Mr

Bending moment strength

Yield moment

My
M

Bending moment resistance

of cross section without normal force of cross section

MN

Bending moment resistance

including

normal force

NOTATION

203

N P

Normal force
Shear retention coefficient
Effective

Pu

design

normal force

on

wall section for calculation of shear stress

Pv

Viscous effect

Q
Re

Transformation matrix

Bending moment resistance in constant elastic region

Nominal minimum

Rm

bending moment resistance

with

zero

normal force

Rp
T{
U

Bending moment resistance in plastic hinge zone

Period of mode i

Global
Global

displacement vector
velocity vector
vector

Global acceleration

Shear force, element volume

Vc VE
VR Vs Vu
Vw
Z

Cracking force in shear, concrete contribution to shear capacity

shear force from static
Shear

equivalent force

strength

Steel contribution to shear capacity Ultimate shear capacity Shear demand

Softening

modulus of confined concrete

Latin Lower Case

ag(t)
be

Ground acceleration

history hoops

Average width of wa cross section

Width of confinement
Wa thickness Lower distance to centre of relative rotation Element damping matrix

bh
bw
c

Static

height

fc

Concrete stress

fc'
fd

Compressive concrete design strength

Damping force

fs
fi

Dynamic force due to ground motion

Eigenfrequency of mode i
Inertia force

fi

204

NOTATION

fs

Equivalent shear stress

Stiffness force

fs
ft

Concrete tense strength

Steel stress due to tension Ultimate
Shear Yield

fts fu
fv

stiffening

strength of steel

strength of concrete

fwy

strength of horizontal waU reinforcement

fy
h

Design yield strength of steel

Macro element height

hc
k

Upper distance

to centre

of relative rotation

Element stiffness matrix Elastic element stiffness matrix

Element stiffness matrix of internal beam

ke kib
Kie

Inelastic element stiffness matrix

kts
l

Concrete tension

softening factor

Macro element length

lc

Centroidal distance
Web

h
m

length

Element mass matrix

Shear magnification factor

ms
n

Number of storeys
WaU aspect ratio

ra
s

Spacing

of reinforcement bars

Sh
t
u u

Spacing of confinement hoops

Web thickness

Geometrical parameter for shear form factor

Element displacement vector

Ui
V

Element velocity vector

Element acceleration vector

Degree

of freedom No. i

Geometrical parameter for shear form factor

Vc
w

Concrete shear

stress

Concrete crack width Concrete crack width at

zero

WQ

tense stress

Xi

Fibre centroidal distance

205

APPENDIX

USER ELEMENT

INPUT DESCRIPTION

The

macro

model

developed in Chapter three

essentiaUy
as

was

coded

as a

"User Element"

[Abaq91],

which may be used the

in the

same manner as

the

library elements.
of this

The function, element

are

input properties,

weU

as some

useful Output

quantities

user

briefly described Only

small

in this

appendix.
are

the features which

a

specific

for the

user

element wl be described here. A

example of

complete input

file is

given

at the end of

Appendix

A. For the

complete use ofthe Software, the reader is referred to the Abaqus

manual

[Abaq91].

Function

The

user

element is coded

as a

subroutine

on a

separate file of source code, and describes

to the discussion in

the nonlinear behaviour of the element each increment, the subroutine

according
an

chapter three.

In

performs

update of the stiffness properties and

a user

the

resulting element forces.

interface, which has
a

These

are

devered to the program via

element subroutine

fixed formal

[Abaq91]. displacements properties

are

In each increment, the element

fed back into the subroutine

one

by

the

program. In order to maintain certain

from

increment to another,

vector

of Solution

dependent

"state variables" is used, which may be

user

changed by the

subroutine
as

and which is stored until later increments. Since the

a

element subroutine is written

separate Fortran file, it has

to

be connected to the program

during execution,

and this is

made by

includmg the foUowing ne in die input file:

*USER SUBROUTINES, L\PUT=15

The

user

element has to be

speeified just

like the

brary

elements. The
corner

simplest
only.

user

element type

describing

the

macro

model is the U30 with four line in die

nodes

It is

speeified by including the foUowing

nes of d.o.f. numbers):

input

file (one line of

speeifications, two

206

APPENDIX A

*USER ELEMENT, TYPE=U30, NODES=4, COORDINATES=2, PROPERTIES=8,

VARIABLES=45

1,2
3, 1, 2, 6

Except what was stated above, the first line specifies that two coordinates are i^uired
per node, the number of

input properties
equal
to

is

equal

to

eight, and the

number of solution-

dependent

State variables is

45. The two last nes

prescribe the

d.o.f. numbers
are

available for the nodes; for the first node the first and second

degrees of freedom

avaable, which remains the

same

until the third node, from which the sixth

degree

of

freedom is also available. It should here be noted that the coordinates of the nodes

represent the location of the end points of the flexural Springs of the

macro

model,

see

chapter three.

Input properties
The

eight input properties prescribed

are

in the ne of element

specification

are

given

on a

line in free format and

described

as

foows (the units

given are recommended but any

other units consistent with other

input data may be used):

1) Cross sectional

area

of entire wa section

(m2)

2) Moment of inertia (about strong axis) of entire wa section (m4) 3) Young's modulus of uncracked
concrete

(MPa)
to

4) Cracking factor for stiffness of flexural Springs in tension, equal

ratio of com

pressive to
0.03)

tense stiffness,

(useful

rnge: 0.4 to

0.8)

5) Yielding factor, equal to ratio of yielded to compressive stiffness (useful rnge: 0.01
to

6) Bending

moment at flexural

yielding of cross

section with

zero

axial load, for

calculation

procedure

see

chapter three (MNm)

see

7) Shear force at the onset of shear cracking, for calculation 8) Cracking factor in shear, equal
calculation
see

chapter three (MN)

to

ratio of cracked to uncracked shear stiffness, for

chapter three, (useful rnge: 0.10 to 0.30).

sectional

A wa element with

a cross

area

of 1.8

m2,

a moment

of inertia of 5.4

m4,

Young's modulus of 33000 MPa, cracking and yielding factors in flexure of 0.7 and
0.03

respectively,

yield

moment

of 6.5 MNm with

a

no

axial load, and

an

uncracked

shear force

capacity

of 2.8 MN and

cracked shear stiffness of 16% of uncracked, will

USER ELEMENT INPUT DESCRIPTION

207

thus
the

require

the

foUowing

user

element property set written

as

follows

on

two lines on

input file, assuming that the

element

belongs to the element set WALL:

*UEL PROPERTY, ELSET=WALL

1.8, 5.4, 33000., 0.7, 0.03, 6.5, 2.8, 0.16

Based upon the information contained in

input properties and

the coordinates ofthe

are

nodes, i.e. the element geometry, essentiaUy all other relevant element properties

computed automaticaUy in die user element subroutine.

Output quantities
For

plotting

purposes,

number of Output

quantities

are

made avaable among the

solution-dependent state variables,

behaviour ofthe
user

which may be useful in the evaluation of the nonnear

are

element. These

written

on an

element file

during execution, for

lines
are

each increment, if within the calculation step, included:

see

[Abaq91],

the

foUowing two

*EL FILE, ELSET=ELOUT,

FREQUENCY=2

SDV

which

assumes

that the element(s) of interest for Output

are

included in the element set

"ELOUT"

(default: all elements), and that information is desired for every second increment).
and these wl in the
case

increment (default: eveiy On

a

"post file",

see

[Abaq91], plot commands may be given

one or more

of a
are

user

element often eontain

nn

of the State variables of this element. These

denoted SDVnn, where

is the number if the

user

specific

State

variable desired. For

instance, the shear force of the

denoted SDV22. As
shear force
vs. a

element is stored in State variable 22, which wl be

brief example, the

input

sequence

on a

post file for

plot

ofthe

time wl here be

given

on

three lines:

*HISTORY TIME BASE SHEAR

SDV22,, SHEAR-EL1, 1, 1,,

The third line

one as

speeifies,

after the State variable number,

as

scaling

factor which is set to

default

(if left blank

here), then

label for the

plot,

then the element number

208

APPENDIX A

(El. No.

one

in this case) and

integration point
see

number

(always

one

for

user

elements).

For closer details and

options,

[Abaq91].
user

Some useful State variables for Output purposes for the

element type 30

are

sted in

foUowing:
SDV22
SDV23 Base shear

(MN)

Shear deformation over element

height (m)

SDV28
SDV29 SDV30 SDV34

Rotational ductity, left curvature

Rotational

ductity, right curvature

% of story

Interstorey (element) drift, Bending moment (MNm)

Curvature

height

SDV35

Example

of

Input File

In order to facitate the

means

understanding
a

of the

use

of the

macro

model, implemented by
is

of a

user

element,

smaU

example of a structural wa modeed by two elements deliberately chosen figure A.1.

presented

here. The small number of elements is

so as to create an

example as simple as possible.

The side view of die wall is shown in

q c

w///////////r+

V///////////7A

6.00

i
Model with element and node numbers
two user

Wall prototype

Figure A.1 Structural wall modelled by

foUowing basic waU properties

elements

oftype

U30

The

are

assumed: Cross sectional

area:

1.8

m2,

cross

sectional uncracked moment of inertia: 5.5

m4, Young's

modulus of concrete: 33000

moment at zero

MPa, flexural cracking factor: 0.7, flexural yielding factor: 0.03, bending
axial load: 6.5 MNm, shear force at

cracking: 2.8 MN, cracked shear stiffness ratio:

load is

0.16. It is further assumed that

gravity

applied

to

the wall at the top

as two

USER ELEMENT INPUT DESCRIPTION

209

concentrated loads each of 0.5 MN, and then

static horizontal force is

apped

to the

wall from the left at the top. The central distance between the two flexural

spring

is

chosen

as

5.5

m.

Furthermore, it is assumed that residual force tolerances

are set as

0.05

MN for forces and 0.1 MNm for moments.

At first, the

input file for

the structural

analysis

is

given

below. On the next page,

post fe is shown for the

same

example containing conunands for plotting

vs.

die second
versus

step ofthe horizontal top displacement

time, the shear force of element one

of element
one.

top

displacement,
*HEADING

and the moment

versus curvature

Example of two story waU

*USER SUBROUTINES, L\PUT=15 *NODE

1, 0., 0. 2, 5.5, 0. 3, 0., 4. 4, 5.5, 4.

5, 0., 8.

6, 5.5, 8.
*USER ELEMENT, TYPE=U30,

NODES=4, COORDINATES=2, PROPER1TES=8,

VARIABLES=45

1,2

3, 1, 2, 6
*UEL PROPERTY, ELSET=WALL

1.8, 5.4, 33000., 0.7, 0.03, 6.5, 2.8, 0.16 ?ELEMENT, TYPE=U30, ELSET=W ALL

1, 1, 2, 3, 4
2, 3, 4, 5, 6
*BOUNDARY

1, ENCASTRE 2, ENCASTRE *STEP, LINEAR

*STATIC *CLOAD

5, 2, -0.5 6, 2, -0.5
*END STEP

210

APPENDDC A

*STEP, INC=20

*STATIC, PTOL=0.05, MTOL=0.1, DIRECT

0.05, 1.0
*CLOAD

5, 1, 1.0
*NODEFILE
U
*EL FILE,

FREQUENCY=1

SDV

*END STEP

Post fe for

use

with element and node data

resulting from calculation with above

input fe:
*POSTFTLE

Example of two story wa

HISTORY, BSTEP=2
TIME DISPL (m)

Ul,, TOP-DISPL, 6, 1,,

VARIABLE-VARIABLE, BSTEP=2
DISPL (m)
SHEAR FORCE (MN)
,

Ul,, TOP-DISPL, 6, 1,

SDV22,, SHEAR-EL1, 1, 1,,

VARIABLE-VARIABLE, BSTEP=2
CURVATURE MOMENT

(MNm)

SDV35,, CURVATURE, 1, 1,,

SDV34,, MOMENT, 1, 1,,

211

APPENDIX

USER MATERIAL INPUT DESCRIPTION

The micro model

developed
well

in

Chapter four was coded

die
same manner as

as a

"User Material"

[Abaq91],

which may be used the

essentiaUy in
as

the

brary materials. The function,

user

input properties,

as some

useful Output

quantities of this

material

are

briefly described in this appendix.

Only
small the features which
a are

specific

for the is

user

material wl be described here. A

example of

complete input file

given

at the end of

Appendix

B. For the

complete use ofthe

Software the reader is referred to the

Abaqus manual [Abaq91].

Function

The

user

material is coded

as a

subroutine in

separate fe of source code, and describes

the nonlinear behaviour of the material at every

integration point according performs

an

to the

discussion in

chapter

four. In each increment the subroutine and the

update

of the

material modulus
via
a user

properties

resulting stresses.
a

These

are

devered to the program

material subroutine interface, which has

fixed format

[Abaq91].
are

In each increment, the strains for each

integration point

fed back into the

increment to

subroutine

by the program.

In order to maintain certain

properties from one

another,

vector of

solution-dependent

"state variables" is used, which may be

user

changed
material

by

the subroutine and which is stored until later increments. Since the
as a

subroutine is written

separate Fortran file, it has

to be

connected to the program

during execution,

and this is made

by including the foUowing ne in the input file:

USER SUBROUTINES, INPUT=15

The

user

material has to be

speeified just like the brary

some

materials. The

user

material is to
as

be used in connection with

or

of the available membrane elements, such

die CPS4,

the CPS8,
way
as

see

the

Abaqus

manual

[Abaq91].

The

user

material has to be named in die

user

same

for

brary materials.

In the material block the

material is

speeified

as

foows:

212

APPENDIX B

USER MATERIAL, CONSTANTS=8

foUowed

by

line with the

eight input constants, described below

in the section

Input

properties. The abovementioned solution-dependent state variables which are 23 for the

simplest user material model are then speeified in the input file as foows:
DEPVAR 23

Input properties
The

eight input properties prescribed

on a

as

constants in the ne of material

specification are
given
are

given

ne in free format and

are

described

as

follows (the units

recommended but any other units consistent with other

input data may be used):

1) Young's modulus of uncracked

concrete to

(MPa), (useful rnge: 20000 to 40000)

2) Poison's ratio (useful rnge: 0.15 3) Strain 4) Strain

at onset at

0.20)

of cracking (useful rnge: 0.00007 to 0.00015)

end of tension

stiffening (useful rnge: 0.0010 to 0.0030)

5) E-modulus for reinforcement steel (MPa) (useful rnge: 180000 to 220000) 6) Yield strength of steel, without regard to tension stiffening (MPa) (useful rnge: 300 to600)
7) Reinforcement ratio in horizontal direction (useful rnge: 0.002
to

0.05)

8) Reinforcement ratio in vertical direction (useful rnge: 0.002 to 0.05)

A material with

concrete E-modulus of 33000

MPa, Poison's ratio of 0.2, cracking

of 0.002, E-modulus for steel of

strain of 0.0001, strain at end of tension

stiffening

200000 MPa, reinforcement ratios of 0.01 and 0.02 in the horizontal and vertical

directions
as

respectively, wl thus requke the foUowing

on one

user

material property set written

foUows

ne

on

the

input fe:

33000., 0.2, 0.0001, 0.002, 200000., 570., 0.01, 0.02

USER MATERIAL INPUT DESCRIPTION

213

Output quantities

For

plotting

purposes,
State

number of Output

quantities

are

made available among the

solution-dependent

variables, which may be useful in the evaluation ofthe nonnear

behaviour of an element

employing the user material. integration point

These

are

written

on an

element fe

during execution, for

each

in the element, and for each increment, if

are

within the calculation step,

see

[Abaq91], the foUowing two lines

included:

EL FTLE, ELSET=ELOUT,

FREQUENCY=2

SDV

which

assumes

that the element(s) of interest for Output

are

included in the element set

"ELOUT" increment

(default:
(default:

a elements), and that information is desired for every second

one

every

increment).

In

"post fe",

see

[Abaq91], plot conunands

one or more

may be

given

and these wl in the

case

of a user element often eontain

of the State variables of this element. In die

is the number if the

post file these

are

denoted SDVnn where

nn

specific

State variable

desired. For instance the stress normal

to

the first crack direction is stored in State

a

variable 9, which wl be denoted SDV9. As

brief

example, given

the

input

sequence

on a

post fe for a plot of the shear force

vs.

time wl here be

on

three lines:

HISTORY
TTME

BASE SHEAR

SDV9,, SHEAR-EL1, 1, 1,,

The third line

one as

speeifies, after the

as

State variable
a

number,

scaling factor which is

set to

default (if left blank

one

here), then

label for the

plot,

then the element number

(El. No.
see

in this

case) and integration point number. For closer details and options,

[Abaq91].
material

Some useful State variables for output purposes for the

in the

user

are

finally sted

foUowing:

214

APPENDIX B

SDV6
SDV7

Accumulated concrete strain normal to first crack direction

Accumulated concrete strain transverse to first crack direction Accumulated shear strain upon

SDV8 SDV9 SDV10

SDV11

cracking

Accumulated concrete stress normal to first crack direction Accumulated concrete stress transverse to first crack direction Accumulated shear stress upon

cracking

SDV15

Accumulated horizontal steel strain Accumulated stress for horizontal steel Accumulated vertical steel strain Accumulated stress for vertical steel

SDV16
SDV18 SDV19

Example

of

Input File

In order to facitate the

means

understanding ofthe
a

use

ofthe

macro

model, implemented by

of a

user

element,

small

example of a structural wa

modeUed

by two elements
so as

will here be
create an

presented.

The small number of elements is

deberately chosen
seen

to

example as simple as possible.

The side view ofthe wa is

in

figure B.1.

10

11

12

5
7
o

6
8
9

3
4

4
5 6

cd

1
1

2
2

V///////Y/Y/.

600

I
Model with element and node numbers

Wall prototype

Figure B.1 Structural wall modelled by six library membrane elements oftype CPS4,
with material behaviour speeified according to the USER MATERIAL

option
(uniform),
strain of

The

following

basic wa

properties

are

assumed: WaU thickness 0.3

Young's modulus of
0.0001, strain
at

concrete:

33000 MPa, Poison's ratio of 0.2,

cracking

end of tension

stiffening

of 0.002, E-modulus for steel of 200000 MPa,

USER MATERIAL INPUT DESCRIPTION

215

reinforcement ratios of 0.01 and 0.02 in the horizontal and vertical directions

respectively. Furthermore,
for forces.

it is assumed that residual force tolerances

are

set to 0.05 MN

The

input file

for the structural

analysis is given

below. On the next page, for

post fe is
of die

shown for the horizontal top

same

example containing commands

vs.

plotting of the second step

displacements

the base shear, and the stress normal to the first crack

direction of element one,

integration point one.

HEADING

Example of two story wa

USER SUBROUTINES, INPUT=15 NODE

1, 0., 0. 2, 3., 0. 3, 6., 4. 4, 0., 2.67 5, 3., 2.67

6, 6., 2.67
7, 0., 5.33

8, 3., 5.33 9, 6., 5.33

10, 0., 8.
11,3.8. 12, 6., 8. ELEMENT, TYPE=CPS4

1, 1, 2, 5,4
2, 2, 3, 6, 5 3, 4, 5, 8, 7

4, 5, 6, 9, 8 5,7,8, 11, 10

6, 8, 9, 12, 11
SOLID SECTION, MATERIAL=WALL

0.3

MATERIAL, NAME=WALL
USER MATERIAL, CONSTANTS=8

33000., 0.2, 0.0001, 0.002, 200000., 570., 0.01, 0.02

216

APPENDIX B

DEPVAR
23

BOUNDARY

1, ENCASTRE

2, ENCASTRE 3, ENCASTRE STEP, LINEAR

STATIC CLOAD

10, 2, -0.5
12, 2, -0.5
END STEP

STEP, INC=20

STATIC, PTOL=0.05, MTOL=0.1, DIRECT

0.05, 1.0
CLOAD

10, 1, 1.0
NODE FILE

U, CF
EL FILE, SDV

FREQUENCY=1

END STEP

Post file for

use

with element and node data

resulting

from calculation with above

input fe:
POSTFILE

Example of two story wa

VARIABLE-VARIABLE, BSTEP=2
DISPL (m) SHEAR FORCE (MN)

Ul,, TOP-DISPL, 12, 1,, CF1,, SHEAR FORCE, 10, 1,,

VARIABLE-VARIABLE, BSTEP=2
N STRAIN
N STRESS

SDV6,, STRAIN, 1, 1,,

SDV9,, STRESS, 1, 1,,

217

APPENDIX

YIELD MOMENT FOR MACRO MODEL

The selection of the

section 3.5. In this

yield

moment for the macro model was described in

chapter three,
storey
on

Appendix, the

selection of die

yield

moment for the three

test

specimen, presented in chapter five, eight storey capacity designed waU.

A moment curvature relation

wl be demonstrated, foowed

by comments

the

was

established for the waU

cross

section, according

to

the

procedure described in chapter three,

a

section 3.5. The results of this

a

procedure apply
waU there are,

rather to the behaviour in

flexural crack. Over

concrete.

certain

height of the

however, large parts of uncracked

Therefore, the average flexural stiffness is

considerably higher than implied by the moment curvature results.

In

figure

Cl the moment curvature relation for the wall

cross
a

section of the test

1:3 scale model of

specimen
for the

used in

chapter five
a

is

given.

The test

specimen,
was

which is

the three lowest stories in

ten

storey buding,

presented
5.4.1,

in section 5.3, and used

reliability

tests of the macro model in section

as

we

as

in

subsequent

reabity tests for die micro

In

model.
as

figure

Cl the moment curvature relation

obtained from the Computer program

described in section 3.5 is

behaviour
across a

displayed.
cracking

This result is in tension may

a

mainly representative
clearly
be

for the of

crack. The

seen at a moment

around 0.5 MNm. From there the program exhibits

relatively soft behaviour due to

now must

the the

fact that essentiaUy a tensile forces

reinforcement. The

over

the

cross

section

be carried

by

yielding behaviour
curve

is

seen

to

develop gradually,

and the onset of

yielding

is from the

estimated to
curve

occur at

around 2.8 MNm. The

gradual yielding

gives

rounded

shape

of die

which continues until around 3.2 to 3.3 MNm where

the behaviour becomes

more

straight,

which

means

that the strain

hardening

alone

dominates the behaviour, rather than the The

yielding of additional bars.

macro

hysteretic model
as

for the outer vertical Unks of die

model

assume a

trilinear

skeleton curve,

shown in
a

chapter three.
This

In

addition, for the uncracked stiffness ofthe

was

compressive prevail
until

branch

cracked tensile branch

oceurs.

assumed. This cracked stiffness wiU

yielding
entire

yielding
a

is for the model assumed to simulate the

manner.

yielding of the
used in

cross

section in

representative

This

means

that

yield

level must be found from the nonnear cracked behaviour in

a

figure 6.1,

which may be

trilinear model.

218

APPENDIX C

5.0

4.0

Behaviour in crack Elastic behaviour Selected model behaviour-

0.000

0.002

0.004 Curvature

0.006

0.008

(rad)
1:3 scale test specimen

Figure Cl

Selection

ofyield moment for

5.0

-ir-

Behaviour in crack
Elastic behaviour Selected model behaviour-

i.

0.000

0.002

0.004

0.006

0.008

Curvature

(rad)

Figure C.2 Flexural behaviour including effect of axial load figure

Cl

In

yield level

of around 3.0 MNm

was

found to be

representative,

as a

single level, representing yielding for the entire

behaviour is denoted in the

cross

section. The selected model

as

figure by

dotted line. The cracked stiffness is taken

around 70 percent ofthe elastic value until the selected

yield level. Thereafter a fraction of

a

yielded

to

uncracked vertical

spring stiffness

is taken in the rnge of

few percent

as

YIELD MOMENT FOR MACRO MODEL

219

discussed in

chapter

three. For the test

specimen

model

yield stiffness of around applied

one

percent

was

found to

give reasonable agreement with experimental dato.

as was on

The model behaviour with the fuU axial load of 0.87 MN,

the test

specimen,

is

seen

in

figure C.2.

The

approximation ofthe

model behaviour
cross

according to

expression

3.38 may here be checked

against the result ofthe only apply

sectional behaviour.

Relatively good agreement may be observed.

It should be noted that

figures

C. 1 and C.2
soon as

to the skeleton

curve

of a the

hysteretic model
of these

of chapter three. As
are no

load reversal has taken

place, the relations

govemed by

figures

longer relevant,

and the flexural behaviour is

unloading and reloading rules as presented in chapter three.

For the

eight storey capacity designed wa, presented dynamic problem repeated

may be

in section 5.3, and used for the the

nonnear

of

chapter five,

the

procedure of determining
pattem
cross as

yield

moment wi not be

since it follows the

same

described above. From

figure 6.4 and 6.6, the moment curvature relation

fll

for the

sections ofthe mited and

ductity design
were

seen.
on

From

there, the yield levels of 6.6 and 3.8 MNm,

with

respectively Increasing

read, based

the

curve

design

values for material

mean

strength.
gives
7.9

these values

by

20 % in order to arrive at effective

values

MNm and 4.6 MNm,

respectively.
6.4 and 6.6 for the
curves

Inspection

of

figures

of

yield strength

of 550 MPa, strain of 30 MPa, weU. The

hardening

of 0.7% of elastic modulus, and concrete

compressive strength

suggests good agreement, and would have given about 7.9 and 4.6 MNm

as

procedure was repeated

for die storeys of the elastic

region.

220

221

APPENDIX

ELEMENTS FOR MICRO MODEL

A short overview of elements avaable for the micro model

developed in chapter four is

to the

presented

in this

Appendix.

The elements shown here

belong

Abaqus

element

brary, [Abaq91].
The

presentation

here is limited to the

plane-stress family of elements,

to

be used in
used

pure two-dimensional problems. The membrane elements could

theoreticaUy also be

for the micro model, in the

case

of three-dimensional

analysis.

CPS3

CPS6

CPS4

CPS4R

CPS8

CPS8R

Figure D.1

Plane stress elements for use with micro model of chapterfiour

222

APPENDIX D

Figure D.1 shows the six available plane-stress elements.

nodes with two

In the

figure the
denote

dots denote

in-plane degrees

of freedom each and the

an

crosses

integration
see

points.

In the element

designation in the figure

R denotes reduced

integration,

e.g.

[Abaq89].
For structural walls the

triangulr elements

may rather be useful in the

case

of

irregularities

or

mesh refinement between

rectangular elements. NormaUy the rectangular

elements will be used, and for the nonlinear models elements CPS4 and CPS8

developed

in this report, the two

proved

to be

preferable.

The elements with reduced

zero
are

integration

CPS4R and CPS8R necessitated finer meshes in order to avoid

may
occur

energy modes which

in

problems

with dominant flexural action, and

they

thus not recommended

for typical structural wa

A

analysis.
three-storey
1:3 scale structural wa test

comparison

was

made between die elements CPS4 and CPS8 for die nonlinear static

monotonic behaviour. The

in

specimen presented
were

chapter three

is used. The meshes A and B,

In

presented

in

chapter five,

used for
same

this

comparison.

figure D.2 is shown the used mesh-element combinations.

The

number of elements used in mesh A with CPS4 (left) and with CPS8
same

(right). VirtuaUy die

number of degrees of freedom

are

used in mesh B with CPS4 (centre) and in mesh

A with CPS8

(right).

>l
**>

--;

S?

*# *

%
""' ^
*

<.

'

<

Figure

D.2 Meshes

for

element

comparison: Mesh

A with CPS4

(left),

mesh B with

CPS4 (centre), mesh A with CPS8

(right)

Figure

D.3 shows the numerical result ofthe element

comparison. The
It is
seen

base shear is

plotted versus the horizontal fourth floor (free end) displacement.

of mesh A with CPS8
It should

that Solution

approaches the Solution of mesh B with CPS4.

as

finaUy be noted that due to the formulation of the micro model

the
same

described in

chapter four,
suggested by

integration scheme is used for both the

concrete and the

reinforcement Different
some

integration

for these materials within

certain element is

authors.

ELEMENTS FOR MICRO MODEL

223

0.01

0.02

0.03

0.04

Fourth floor horizontal

displacement [m]

Figure D.3

Base shear

versus

fourth floor horizontal displacement: comparison between

element CPS4 and CPS8

224

225

APPENDIX

FREQUENCY STUDY OF DAMAGED STATES

In order to

display the frequency

examples
on

behaviour of the

gradually damaged eight storey wall chapters

five and six,
are
a

used in the

the nonnear

dynamic

behaviour in

few

free Vibration studies below.

The
macro

were

performed.

The results in form of eigenfrequencies

given

model aUowed for the

macro

possibility

of

changing the stiffness of the internal

was

Springs of

chosen

elements. This feature

used in order to simulate free

Vibration behaviour in various simulated in this

manner.

damaged Springs

states. Four different states of

as

damage

were

The first state, referred to of all

completely elastic
elements have
an

State, simulates

the behaviour for which aU four

macro

elastic stiffness.

The second State, the cracked State, is achieved

macro

by letting

the vertical outer

spring of each
acr of 0.7
was

element

develop

cracked stiffness. In this

case a

cracking factor

used,

as

described in

chapter three.

The third State simulates moderate flexural

yielding in equal

the three

plastic hinge

elements (see mesh,

figure 5.20).

This is achieved

by setting the
to

stiffness of the outer vertical

Springs

at one side of the

was

plastic hinge
as

elements

yielded

stiffness. The
a

yield

factor used here

State in which

taken

0.01. The fourth stete is and is obtained

intended to simulate

damaged

major yielding occurred,

by
all

letting the
(i.e. close

central vertical
to

spring

ofthe

plastic hinge elements develop yielded stiffness

are

zero). The eigenfrequencies for the eight first modes, which

are

horizontal modes,

given

on

die

following page.
obtained

For the micro model the elastic free Vibration behaviour

was

by prescribing
which starts

high

tense

strength

and

high yield

limit. Since

cracking

is

procedure

gradually

and needs

but-up

stress states to form

properly

it

was not

found feasible to
same

perform a meaningful cracked State study by means

essentiaUy to
a

ofthe micro model. The

appes
eight

yielded State.
by
means

The elastic State

only wl therefore be displayed compared

for the

lowest modes
state obtained

of the micro model. This State may be

to the elastic

by the macro model. The comparison of the two elastic damaged states, simulated by the
that the
macro

states is shown in

table E.l. The

model,

are

shown in table E.2.

From table E. 1 it is

seen

frequencies for the lower modes

are

in

fairly good
model. lowest

agreement with both models. The first mode is about 6 % lower for the
Somewhat refined meshes for the micro model devered about the

macro
same

frequencies

as

for the

macro

model.

226

APPENDDC E

Mode Number
1
2

f (Hz) Macro Model 0.74

f (Hz) Micro Model

0.79

4.25
10.4 17.4

4.60
11.3 18.4

3
4

5 6
7 8

24.3
29.9 34.2
37.0

23.3

24.5
29.1

32.3

Table E.l

Eigenfrequencies for elastic behaviour

Mode Number
1
2

Cracked State 0.67

3.96 9.79

Moderate Yielding 0.54 3.90

9.90 16.7
23.4

Major Damage
0.19 2.97 7.16 13.9 21.5 28.3 33.5

3
4

16.7 23.5

5 6
7 8

29.3
33.6

29.3 33.9

36.5

36.9

36.8

Table E.2

Eigenfrequencies (Hz) obtained by

macro

model for

eight storey

wall

of

chapter five, for different states of damage

As for the
seen

damaged

states, obtained

by the macro model,

the

tendency

may be

clearly
are

that the lowest modes shift towards smaUer

frequencies
were

while ie

higher modes setting

unaffected. It should be

kept

in mind that these states

achieved by

certain

stiffnesses to fixed values.

During the time history analyses, however, the stiffhesses are

constantiy changing.

The

damage

State

behaviour

suggested here
be

is therefore

only

simpfication.
oceurring e.g.

The

"major damage"
a

State may not

appcable

to any free

Vibration

after

strong ground motion, since certain parts ofthe

structure wiU then to

alternatingly
yielding
bound
in
on

describe

unloading behaviour
complement

with

typical larger stiffnesses compared

seen as a

loading. Therefore,
as a

this State should rather be

hypothetical

lower

"free Vibration",

to the upper bound of elastic

behaviour. It

FREQUENCY STUDY
should be noted that

OF DAMAGED STATES

227

during

time

history analysis,
as

cracked flexural stiffness is assumed

already
As
a

at the

beginning ofthe analysis,

to

described in chapter three.

comparison

the different states of damage estimated

may be

by the

free Vibration data,

the behaviour
In

during the time history analysis

to E.3 the Fourier

displayed

in the

frequency region.
three different

figures E.l

spectrum ofthe base shear ofthe limited ductility design

of the

eight storey

wall of chapter five is shown. The three

plots represent

time segments.
which is

Figure

E.l shows the spectrum for the time segment 0 to 4 seconds,

largely dominated by elastic/cracked behaviour. Figure E.2 shows the spectrum

for the time segment 4 to 8 seconds, which contains substantial

yielding

behaviour.

Figure E.3, finally, shows

which is

the spectrum obtained for the time segment 8 to 12 seconds,

largely dominated by free Vibration behaviour. eigenfrequencies

with the
may be

The movement of the lower

seen

from the spectra, and

shows

relatively good agreement

E.l

suggestions

from free Vibration studies of may be

damaged
figure

states. For 0 to 4 seconds the first three

as:

frequencies

clearly

seen

in be

0.7 Hz, 4.3 Hz, and 9.5 Hz. These values should

essentially

comparable
seen

to a mixture of the free

Vibration values for elastic and cracked states. It is

that this is

largely the

case.

For time 4 to 8 seconds the

figure

E.2 shows the three first

to

frequencies essentially

as:

0.5 Hz,

4.1 Hz, and 9.2 Hz. These values

state of free Vibration.

seem

correspond largely to the "moderate yielding"

we

Lastly, for time

8 to 12 seconds,

have

damaged
in

system in free Vibration. The first mode dominates the damping out ofthe vibrations. It is
here
seen

that the time

history analysis ended at

point

at which the

system

was

Figure

E.l

Base shear spectrum

of eight storey wall, time segment 0

to 4

seconds,

characterised as elastic behaviour

228

APPENDIX E

Figure

E.2

Base shear spectrum

of eight storey wall,

time segment 4 to 8 seconds,

characterised as moderate yielding behaviour

Figure

E.3

Base shear spectrum

of eight storey wall,

time segment 8 to 12 seconds,

characterised mainly asfree Vibration

damaged

State with the lowest

are

frequencies

as

follows: 0.5 Hz, 3.8 Hz and around 9 Hz.

These values

placed in between the

are

"moderate

yielding"

and

"major damage"
yielding State.

states of

the free Vibration tests, but

only

little lower than the moderate

This confirms the earer mentioned Statement, that the

major damage

State should

normally only be regarded

as a

lower bound. In the here shown case, the wall may be

FREQUENCY STUDY OF DAMAGED STATES

229

regarded
moderate

as

being capable
some

of

withstanding

the

ground

motion

suggested

here with

yielding, and kept in

local

major yielding. ground

acceleration wl affect the extent
are

It should be

mind that die level ofthe The

degree the damage wl reach.

data used here and

findings of this study

therefore limited to the

input

assumptions of modelling.

230

231

APPENDIX

DESIGN

DEFINITIONS

In the

presentation of chapters
such

five and six,

are

number of definitions
are

mainly related to

the

capacity design of structural was

the

used. These definitions

A

discussed in deta in discussion of these

design references,

as

[PBM90].

comprehensive

definitions could in

general not be made in this study.

In order to facitate the

understanding of die definitions used in this study,

The summary is
are

summary

is

provided

in this

Appendix.

given as

listing,

where

expressions

and
are

relations between the various definitions

stated

briefly.

The definitions sted here

essentially

mited to the

subject of

structural walls, rather than to the wider

area

of

general capacity design.

Equivalent
The

Static Force Method

according by

to the Swiss Standard SIA 160

earthquake action

may be modeUed

an

equivalent static force.

in deta,
see

Section 4 19 of

the Swiss Standard SIA 160 treats this

modelling

[SIA 160]. Further

the most

explanations

of die clauses of die Standard SIA 160

are

given in [Bach89]. Only

are

important definitions pertaining to the equivalent static force calculation

The fundamental

given below.
forces

frequency

for

buildings

where the horizontal

may

earthquake

are

carried

by

structural was and

cores or as

diagonal bracing,

according

to Clause 419

64 ofthe SIA 160 be estimated

foows

41
h

where

/0

:
:

fundamental

frequency of die buding in Hz

-

Cs

ground coefficient:

stiff ground medium-stiff ground

:
:

Cs Cs

0.9 to 1.1 0.7 to 0.9 under

dimension in

of the

buding in the clirection of Vibration

consideration
h
:

height in

of the

buding

measured from the level of embedment

232

APPENDIX F

The static

equivalent

force

according to

Clause 419 506 ofthe SIA 160 is defined

as

follows

Qacc^Cfa+^VaccQr)
o

where

Qm^m
ah

total horizontal

equivalent force figure

27

horizontal acceleration in accordance with Clause 4 19 62 and

ofthe SIA 160 (or figure 5.26b of this

mental

report)

as

function ofthe funda

frequency ofthe structure which may be obtained according to

Clause 4 19 64
g
:
:

acceleration due to

gravity

Ck

construction coefficient in accordance with Clause 419 71 which takes into account
tions and the
a

reduetion of the equivalent force due to

of design values. It is calculated
as

plastic deforma

use

Ck=Cd
*

where

deformation coefficient
account the

according

to

Clause 4 19 72, taking into

influence of plastic deformations.

According to the
on

SIA

160, the deformation coefficient depends

structural classes
a

(Clause

4 19

74) for which structural wall buildings obtain

deformation coefficient of 2.0, 1.7, and 1.3 for structural classes

SCI, SCII, and SCHI, respectively. The capacity design method

[PBM90] allows

replacement

of the deformation value with

a as

global displacement ductility

follows. Restricted

factor pa which may be chosen

=

ductity:

pa
to

3, fu ductility: pa

5.

Cd

design coefficient according

account the

Clause 4 19 73, which takes into

difference between design values (minimum values of

strength, reduetion of the resistance factor, reduetion of failure

strain) and probable values in earthquake conditions. The design
coefficient is taken
The
as

Cd

0.65.
be
seen as an

design coefficient

can

overstrength

reduetion

factor

[Bach91], [Bach93], obtained

as

DESIGN DEFINITIONS

233

Cd= t
where

=0.65
1.25 1.2 1.5

1-5
=

fo

l0yR

Gm

: mean

value of the

self-weight of the load-bearing structure, according to

Clause 3 22 ofthe SIA 160

X VaccQr

'sum

0I* fc accompanymg actions

oceurring simultaneously with the

3 25 3 ofthe SIA 160

earthquake in accordance with Clause

Yacc
'

load factor for

accompanying action to earthquake as predommant

as

action, usuay taken

0.3, according to Clause 3 24 ofthe SIA 160

Qr

representative value of action, according to Clause 3

24 ofthe SIA 160

For budings, the horizontal force shall be distributed in the

over

the

height of the

structure

foUowing manner, according to Clause 4

19 508

=n

K+v^&),A
Y\G*+yLVaccQr).h

where

Qm.ci

:
'

component of the horizontal equivalent force Q,..

vertical action at storey i

at

height h,

{Gm + ^ YaccOr).
ht

height measured from the level of embedment, as

Clause 4 19 507

discussed in

Cross sectional Actions

ME

bending moment resulting from horizontal equivalent earthquake force

components

^.Cil

distributed

according to Clause 419 508, described

case

above, of the equivalent static force calculation (in the

redistribution in Systems of structural walls,

of moment

ME

is the redistributed

moment)

234

APPENDIX F

MIMC
v

M0

X0MR

0>o,JdE

_2_MR > yRME

1.2 1.0 4_E

YrMe
--ME

Figure F.l

Levels

offlexural strength

Nr.

effective normal force actmg

on

the stractural wall, dueto permanent

as

gravity loads tributary to waU.

the
area

Calculated

\Gm + ^ WaccQr) acting on

tributary to

the waU.

shear force
nents

resulting from horizontal equivalent earthquake force compo

distributed according to Clause 4 19 508, described above, of

Qm^i

die

equivalent static force calculation

Cross sectional

strength according to SIA 162, usuay taken

1.2.

Yr

resistance factor
accounts
-

as

for:

Deviation of real structural system

compared to calculated

Simplifications and inaccuracies ofthe strength model

Cross sectional inaccuracies

M:

flexural demand, calculated

as

yRME

and vahd for a normal force

NG

(note:

not

yR NG)

MB

avaable flexural material

strength,

calculated with

design value (nominal)

not

strengths. Including effect of normal force NG (note:

Must be

yRNG.).

equal to,

or

exceed the flexural demand.

DESIGN DEFINITIONS

235

M0

flexural

overstrength,
is the
as

calculated

as

XBMR

or as

^>0BIME

where

X0

overstrength factor for reinforcement steel, generally

strength" given below,

calculated

shown under section "Material

as

usually obtained

1.2

or more

o,*
Va

'

overstrength factor for wa cross sectional, calculated as M01ME

shear demand

(equal to design shear force),

calculated

as

coJbotttVE

(ov

dynamic magnification factor for shear, calculated as

for
for

buildings

up to six stories

cov
a),
'

0.9

buildings over 6

six stories

=1.3

<

1.8

30

VR

shear

strength at design level, calculated with design material strengths

=

asVR Vc

Vc

V.

concrete

contribution to shear strength, expressed

as

Vc

v.bjd

where vc is the concrete shear stress

given

as

where

Pu

is the effective normal force

acting

on

the

cross

section and
as

Ag

is the gross area,

bw

is the wall thickness, and d may be taken

0.8

times the wa

length Lw.

Vs

reinforcement contribution to shear

where

strength, expressed

as

V.

\fy

A,

is the

area

of shear reinforcement

(horizontal) bars in both

faces, f is the design yield strength, d is the distance described above,

and
s

is the vertical

spacing between the bars.

Material

strength
:

design value (nominal value) for steel yield strength (design strength) overstrength value for steel yield strength, calculated

fo

as

Xjy

236

APPENDDC F

fs'fy k
fo-,
f^f
-

Kf

'oJy

J m-y

TmJy

(U
Figure F.2 Levels of material strength

X0

overstrength factor for reinforcement steel, generaUy calculated as

f-y
Jy

usuaUy obtained

as

1.2

or more

e(f0 y)
/

steel strain at

overstrength, usuaUy taken in the rnge of 2% to 4%

value for steel yield

m.y

effective

mean

strength

material
Jm,

strength reduetion factor for steel, calculated

as

Jm,y

nt

Jy

usuaUy obtained

as

1.2

fcw.min

'

nrtimum value for test-cube

SIA 162

compressive concrete strength according to

(3 23)

fc

design value (nominal value) for compressive concrete strength,

(design strength) obtained
as

0.65 fcw,nUn