FI
RO
UGO
ILIPE PIN
ODRIGU
NHEIRO
UES
Un
An
Co
ar
Bi
co
iversidade d
o 2012
omporta
rmado
iaxial sei
olumns
de Aveiro D
mento s
ismic be
Departamento
smico b
ehaviour
o de Engenh
biaxial de
of reinfo
haria Civil
e pilares
orced co
de beto
ncrete
o
HU
FI
RO
UGO
ILIPE PIN
ODRIGU
NHEIRO
UES
Un
An
Co
ar
Dis
req
rea
Va
Civ
Dia
Fa
Ap
Un
bo
BD
niversidade
no 2012
omporta
rmado
ssertao ap
quisitos nece
alizada sob a
arum, Profess
vil da Univer
as Arde, Pr
aculdade de E
poio financeir
niversidade d
lsa de Douto
D/REITORIA/
de Aveiro D
amento s
presentada
essrios ob
a orientao
sor Associad
rsidade de Av
rofessor Asso
Engenharia d
ro da Reitoria
de Aveiro atr
oramento
/4/47/2008
Departament
ssmico b
Universidad
bteno do g
cientfica do
do com Agre
veiro e coori
ociado do De
da Universid
a da
avs da
to de Engenh
biaxial de
de de Aveiro
grau de Douto
o Doutor Hum
gao do De
entao do D
epartamento
ade do Porto
Apoio finan
mbito do Q
4.1  Forma
comparticip
Europeu e
MCTES atr
doutoramen
haria Civil
e pilares
para cumpr
tor em Engen
mberto Salaz
epartamento
Doutor Antn
o de Engenha
o.
nceiro da FCT
QREN  POP
ao Avana
pado pelo Fu
por fundos n
ravs da bols
nto SFRH/BD
s de bet
imento dos
nharia Civil,
zar Amorim
de Engenha
nio Jos Coe
aria Civil da
T e do FSE n
PH  Tipologi
ada,
undo Social
nacionais do
sa de
D/63032/200
o
aria
elho
no
a
09
HU
FI
RO
UGO
ILIPE PIN
ODRIGU
NHEIRO
UES
Un
An
Bi
Th
the
su
Ag
co
the
Fin
Un
gra
niversidade
no 2012
iaxial seis
esis submitt
e degree of D
pervision of
ggregation at
supervision
e Faculty of E
nancial supp
niversity of Av
ant BD/REIT
de Aveiro D
smic beh
ted to the Un
Doctor of Phi
Dr. Humbert
t the Civil En
of Dr. Antn
Engineering
ort provided
Aveiro through
TORIA/4/47/2
Departament
aviour of
niversity of Av
ilosophy in C
to Salazar Am
gineering De
nio Jos Coe
of the Unive
by the
h the PhD
2008
to de Engenh
f reinforce
veiro to fulfil
Civil Enginee
morim Varum
epartment of
elho Dias Ar
rsity of Porto
Financial su
and FSE in
POPH  Typ
Training, su
Social Fund
MCTES thr
SFRH/BD/6
haria Civil
ed concre
the requirem
ring, under t
m, Associate
f the Univers
de, Associa
o.
upport provid
n the Framew
pologies 4.1
upported by
d and Nation
rough the Ph
63032/2009
ete colum
ments to obta
he scientific
Professor w
ity of Aveiro
te Professor
ded by the F
work of QREN
 Advanced
the Europea
nal funds of t
hD grant
mns
ain
with
and
r at
CT
N 
an
he
...
O Mundo no se fez para pensarmos nele
(Pensar estar doente dos olhos)
Mas para olharmos para ele e estarmos de acordo...
Eu no tenho filosofia: tenho sentidos...
Se falo na Natureza no porque saiba o que ela ,
Mas porque a amo, e amoa por isso,
Porque quem ama nunca sabe o que ama
Nem sabe por que ama, nem o que amar ...
Amar a eterna inocncia,
E a nica inocncia no pensar...
In O Guardador de Rebanho
Alberto Caeiro
o jri
presidente Prof. Doutor Joo de Lemos Pinto
Professor catedrtico da Universidade de Aveiro
vogais Prof. Doutor Jos Manuel Jara Guerrero
Professor Titular da Universidad Michoacana San Nicols de Hidalgo, Morelia, Mxico
Prof. Doutor Anbal Guimares da Costa
Professor Catedrtico da Universidade de Aveiro
Prof. Doutor Humberto Salazar Amorim Varum
Professor Associado com Agregao da Universidade de Aveiro
Prof. Doutor Antnio Jos Dias Arde
Professor Associado da Faculdade Engenharia da Universidade do Porto
Prof. Doutor Alfredo Peres de Noronha Campos Costa
Investigador Principal do Laboratrio Nacional de Engenharia Civil
Prof. Doutor Rui Jorge Silva Moura Pinho
Professor Auxiliar da Universit degli Studi di Pavia, Itlia
Prof. Doutor Nelson Saraiva Vila Pouca
Professor Auxiliar da Faculdade de Engenharia da Universidade do Porto
Prof. Doutor Patrcio Antnio Almeida Rocha
Professor Adjunto da Escola Superior de Tecnologia e Gesto do Instituto Politcnico de Viana do
Castelo
acknowledgements
The present work has been developed at the Civil Engineering Department of
Aveiro University and at the LESE, Laboratory of Earthquake and Structural
Engineering laboratory of the Faculty of Engineering of the University of Porto.
I wish to express my sincere gratitude to the following persons, which have
contributed to make this work possible and helped me growing.
I would like to express my gratitude to my advisor, Professor Humberto Varum,
for his support and guidance throughout the research. Along these years he
stimulated not to be only a PhD student but to go further, giving me the
opportunity to take part in many other activities.
To Professor Antnio Arde, my cosupervisor I would like to express my
appreciation for his precious advices, trust and tremendous support in the
development of the experimental campaign.
To Professor Anbal Costa for his support and contributions along the
development of this work. Since the beginning he was enthusiastic with this
project in particular inciting me for the improvement of the simplified model.
I would like to thank to Xavier Romo for the interesting discussions, support
and information supplied, to Pedro Delgado and Patrcio Rocha, for the
knowledge and information given from previous experiments, and to Romeu
Vicente for the long talks and continuous encouragement.
To Professor Antnio Gil Andrade Campos I would like to express my sincere
gratitude for all the support and assistance in the application of the optimization
methodologies.
I should emphasize that the experimental work described in this thesis would
not have been possible without the assistance of LESE laboratory personnel, in
particularly Valdemar Lus, Andr Martins and Lus Noites. Without them, these
experiments would not be pleasant.
To all my colleagues from the PhD room in University of Aveiro, in particular to
Helena Paiva, Catarina Fernandes and Carlos Couto, I would like to thank for
their patience supporting my moods and my loud music over this years.
I am also grateful to my friends, for their support, patience and for continuing
being my friends despite frequently missing my company. I would also like
express my thanks to my grandparents, parents and brother for all their love
and support throughout my life.
Finally, I would like to dedicate this thesis to Marisa and Maria Carolina, for
their support, love and for always being my core.
palavraschave
Pilares de beto armado, comportamento cclico, ensaios experimentais
biaxiais, modelos nolineares refinados, modelos nolineares simplificados
resumo
A anlise dos efeitos dos sismos mostra que a investigao em engenharia
ssmica deve dar especial ateno avaliao da vulnerabilidade das
construes existentes, frequentemente desprovidas de adequada resistncia
ssmica tal como acontece em edifcios de beto armado (BA) de muitas
cidades em pases do sul da Europa, entre os quais Portugal. Sendo os pilares
elementos estruturais fundamentais na resistncia ssmica dos edifcios, deve
ser dada especial ateno sua resposta sob aes cclicas. Acresce que o
sismo um tipo de ao cujos efeitos nos edifcios exige a considerao de
duas componentes horizontais, o que tem exigncias mais severas nos pilares
comparativamente ao unidirecional.
Assim, esta tese centrase na avaliao da resposta estrutural de pilares de
beto armado sujeitos a aes cclicas horizontais biaxiais, em trs linhas
principais.
Em primeiro lugar desenvolveuse uma campanha de ensaios para o estudo
do comportamento cclico uniaxial e biaxial de pilares de beto armado com
esforo axial constante. Para tal foram construdas quatro sries de pilares
retangulares de beto armado (24 no total) com diferentes caractersticas
geomtricas e quantidades de armadura longitudinal, tendo os pilares sido
ensaiados para diferentes histrias de carga. Os resultados experimentais
obtidos so analisados e discutidos dando particular ateno evoluo do
dano, degradao de rigidez e resistncia com o aumento das exigncias de
deformao, energia dissipada, ao amortecimento viscoso equivalente; por
fim proposto um ndice de dano para pilares solicitados biaxialmente.
De seguida foram aplicadas diferentes estratgias de modelao nolinear
para a representao do comportamento biaxial dos pilares ensaiados,
considerando nolinearidade distribuda ao longo dos elementos ou
concentrada nas extremidades dos mesmos. Os resultados obtidos com as
vrias estratgias de modelao demonstraram representar adequadamente a
resposta em termos das curvas envolventes foradeslocamento, mas foram
encontradas algumas dificuldades na representao da degradao de
resistncia e na evoluo da energia dissipada.
Por fim, proposto um modelo global para a representao do comportamento
nolinear em flexo de elementos de beto armado sujeitos a aes biaxiais
cclicas. Este modelo tem por base um modelo uniaxial conhecido, combinado
com uma funo de interao desenvolvida com base no modelo de Bouc
Wen. Esta funo de interao foi calibrada com recurso a tcnicas de
otimizao e usando resultados de uma srie de anlises numricas com um
modelo refinado. ainda demonstrada a capacidade do modelo simplificado
em reproduzir os resultados experimentais de ensaios biaxiais de pilares.
keywords
RC columns, cyclic behaviour, biaxial experimental testing, refined nonlinear
modelling, simplified nonlinear modelling
abstract
Recent earthquakes around the world have shown that earthquake engineering
research should focus on the vulnerability assessment of existing
constructions. Quite often these constructions are lacking adequate seismic
resistance as in the case of several reinforced concrete buildings.
Since the columns are key structural elements for the adequate seismic
performance of buildings, special attention should be given to their structural
response under load reversals. Moreover, earthquake effects generally require
the inclusion of two horizontal component loads that are recognized to be more
damaging than onedirection actions.
The present thesis focuses on the assessment of the structural response of RC
columns under bidirectional horizontal loads in three main streamlines.
First, an experimental testing campaign was performed on 24 rectangular
building columns, for different types of loading. Two specimens of each column
crosssection type were uniaxially tested, one in each direction (strong and
weak). All the other specimens were tested under bidirectional loading
conditions for different paths. All columns were tested under constant axial
loading conditions. The experimental results are presented and the global
behaviour of the tested columns is discussed, particularly focusing on the
damage evolution, stiffness and strength degradation associated to the
increasing demands, energy dissipation and equivalent viscous damping. In
this framework, one proposal is introduced for a biaxial damage index and
validated against the experimental results.
Subsequently, the tested columns were simulated with different nonlinear
modelling strategies. The studied models are classified into two categories,
according to the nonlinearity distribution assumed in the elements: lumped
plasticity and distributed inelasticity. The analyses show that the global
envelope response is satisfactorily represented with different modelling
strategies, but significant differences were found in terms of strength
degradation for higher drift demands and of energy dissipation.
Finally, a simplified hysteretic model is proposed for the representation of the
nonlinear response of reinforced concrete members subjected to biaxial
bending combined with constant axial load. The proposed model corresponds
to an upgrade of an existing uniaxial hysteretic model, with piecewise linear
behaviour, and adopts an interaction function based on the formulation of
BoucWen smooth hysteretic model. The proposed biaxial model requires the
same type of information as for the corresponding uniaxial one, along with a
correcting term given by an interaction function which modifies the response in
each uniaxial direction in order to couple the two directions responses. For the
calibration of the proposed interaction function, optimization techniques were
used in order to adjust the required parameters. The validity of the simplified
model is demonstrated through the simulation of the response of reinforced
concrete columns tested under biaxial loading.
Table of Contents
TABLE OF CONTENTS ......................................................................................................... I
LIST OF FIGURES .................................................................................................................. V
LIST OF TABLES ................................................................................................................ XIII
CHAPTER 1  INTRODUCTION ....................................................................................... 1
1.1 GENERAL ......................................................................................................................... 1
1.2 RESEARCH PROGRAMME OBJECTIVES ........................................................................... 4
1.3 THESIS SCOPE AND ORGANIZATION ............................................................................... 4
CHAPTER 2  REVIOUS RESEARCH AND BACKGROUND .............................. 7
2.1 INTRODUCTION ............................................................................................................... 7
2.1.1 Typical causes and consequences of column failure during earthquakes ........... 8
2.1.2 Hysteretic behaviour of RC columns ................................................................ 12
2.2 EXPERIMENTAL STUDIES ON THE SEISMIC BEHAVIOUR OF RC COLUMNS ................. 15
2.2.1 Introduction ...................................................................................................... 15
2.2.2 Specimen geometries ......................................................................................... 18
2.2.3 Displacement patterns ...................................................................................... 19
2.2.4 Behaviour of RC columns under biaxial bending with constant axial load ...... 21
2.2.5 Behaviour of RC columns under biaxial bending with varying axial load ....... 25
2.2.6 Biaxial tests on global bare frame structures .................................................... 26
2.3 NUMERICAL MODELLING OF BIAXIAL BENDING OF RC COLUMNS ............................. 29
2.3.1 Introduction ...................................................................................................... 29
2.3.2 Numerical modelling strategies at the element level ......................................... 30
2.3.3 Modelling biaxial flexure with axial force ......................................................... 31
2.4 FINAL REMARKS ........................................................................................................... 34
CHAPTER 3  TEST PROGRAM ..................................................................................... 37
3.1 INTRODUCTION AND OBJECTIVES ................................................................................ 37
ii
3.2 SPECIMENS AND CONSTRUCTION ................................................................................. 38
3.2.1 Introduction ...................................................................................................... 38
3.2.2 Prototype columns: geometry and section detailing ......................................... 39
3.2.3 Construction ..................................................................................................... 40
3.3 MATERIAL PROPERTIES ................................................................................................ 41
3.3.1 Concrete............................................................................................................ 42
3.3.2 Reinforcement steel .......................................................................................... 43
3.4 AXIAL LOAD ................................................................................................................... 44
3.5 HORIZONTAL DISPLACEMENT PATHS ........................................................................... 45
3.6 TEST SETUP AND INSTRUMENTATIONS ....................................................................... 47
3.6.1 Test setup ........................................................................................................ 47
3.6.2 Instrumentation ................................................................................................ 49
CHAPTER 4  ANALYSIS OF EXPERIMENTAL TEST RESULTS .................. 53
4.1 GENERAL OVERVIEW .................................................................................................... 53
4.2 DAMAGE EVOLUTION IN THE TESTED COLUMNS ......................................................... 54
4.2.1 Observed damage ............................................................................................. 54
4.2.2 Damage distribution ......................................................................................... 57
4.2.3 Damage versus drift demand ............................................................................ 59
4.2.4 Performance objective ...................................................................................... 62
4.3 FORCEDISPLACEMENT HYSTERETIC BEHAVIOUR ....................................................... 64
4.3.1 Global analysis ................................................................................................. 64
4.3.2 Evaluation of the yielding displacement ........................................................... 72
4.3.3 Influence of the biaxial loading path on the columns ultimate ductility ........ 74
4.3.4 Coupling effect in the columns biaxial response .............................................. 75
4.3.5 Correlation between maximum strength and yield strength ............................ 77
4.4 STRENGTH DEGRADATION ............................................................................................ 77
4.5 STIFFNESS DEGRADATION ............................................................................................ 80
4.6 DISSIPATED ENERGY ..................................................................................................... 83
4.6.1 Cumulative dissipated energy ........................................................................... 83
4.6.2 Individual cycle energy ..................................................................................... 85
4.6.3 Total dissipated energy until conventional collapse ......................................... 87
4.6.4 Normalized dissipated energy vs displacement ductility .................................. 89
4.7 EQUIVALENT VISCOUS DAMPING RATIO ....................................................................... 91
4.7.1 Evaluation of equivalent damping from experimental results .......................... 91
4.7.2 Empirical proposals for equivalent damping in RC elements under uniaxial
loadings ........................................................................................................................ 97
4.7.3 Equivalent biaxial damping .............................................................................. 99
4.8 ULTIMATE DISPLACEMENT CAPACITY ....................................................................... 101
4.9 DAMAGE QUANTIFICATION ......................................................................................... 105
4.9.1 Original Park and Ang damage index ............................................................ 105
iii
4.9.2 Application of Park and Ang damage index to the uniaxial tests .................. 107
4.9.3 Proposals of damage index for RC columns under biaxial loading conditions 109
4.9.4 Results ............................................................................................................. 110
4.10 FINAL REMARKS .......................................................................................................... 113
CHAPTER 5  NONLINEAR ANALYSIS OF RC COLUMNS SUBJECTED
TO BIAXIAL LOADS .......................................................................................................... 117
5.1 INTRODUCTION AND OBJECTIVES .............................................................................. 117
5.2 NUMERICAL TOOL AND MODELLING STRATEGIES ..................................................... 118
5.3 ELEMENT MODELLING STRATEGIES ........................................................................... 118
5.4 MATERIALS PROPERTIES ............................................................................................ 120
5.4.1 Concrete .......................................................................................................... 120
5.4.2 Reinforcement steel ......................................................................................... 120
5.5 COMPARISON BETWEEN MODELLING STRATEGIES ................................................... 121
5.5.1 Shear drift envelopes ....................................................................................... 121
5.5.2 Cyclic response ................................................................................................ 127
5.6 FINAL REMARKS .......................................................................................................... 133
CHAPTER 6  SIMPLIFIED MODEL FOR THE NONLINEAR
BEHAVIOUR OF RC COLUMNS UNDER BIAXIAL BENDING .................... 135
6.1 INTRODUCTION ........................................................................................................... 135
6.2 UNIAXIAL HYSTERETIC MODEL .................................................................................. 136
6.3 BIAXIAL BOUCWEN MODEL ...................................................................................... 138
6.3.1 Background ..................................................................................................... 138
6.3.2 Framework of the proposed biaxial model ...................................................... 141
6.4 PARAMETER IDENTIFICATION FOR THE SCALING INTERACTION FACTOR ............... 142
6.4.1 Optimization method ...................................................................................... 142
6.4.2 Optimization of the interaction function ........................................................ 144
6.5 VALIDATION OF THE MODEL WITH RESULTS FROM CYCLIC TESTS .......................... 149
6.5.1 Introduction .................................................................................................... 149
6.5.2 Analysis of the results ..................................................................................... 149
6.6 FINAL COMMENTS ....................................................................................................... 154
CHAPTER 7  CONCLUSIONS AND FUTURE RESEARCH ............................ 157
7.1 SUMMARY OF CONCLUSIONS ....................................................................................... 158
7.2 FUTURE DEVELOPMENTS............................................................................................ 160
APPENDIX A EXPERIMENTAL RESULTS ......................................................... A.1
APPENDIX B NUMERICAL SIMULATIONS ...................................................... B.1
APPENDIX C  PUBLICATIONS .................................................................................. C.1
List of Figures
Figure 1  Flexural deficient behaviour in RC columns: a) in the San Salvatore hospital, in
the 2009 LAquila Earthquake; b) rectangular column in residential building in the
2011 earthquake in Lorca, Spain .................................................................................... 8
Figure 2  Shear deficient behaviour in RC columns in residential buildings in the 2009
LAquila Earthquake: a) square column [26]; b) circular column [27]; failure is related
to the transversal reinforcing steel; c) deficient flexural/shear behaviour in a corner
column in a residential building in the 2011 earthquake in Lorca, Spain ..................... 10
Figure 3  Inadequate joint behaviour in residential buildings: a) inappropriate anchorage of
beam longitudinal reinforcement b) deficient detailing of a beamcolumn joint in a
residential building (2009 earthquake in LAquila, Italy); c) RC building collapse in
the 2011 earthquake in Lorca, Spain ............................................................................. 11
Figure 4  Column failure due the presence of nonstructural elements. Short column effect
caused by window openings in residential building: a) 2009 earthquake in LAquila,
Italy; and b) 2011 earthquake in Lorca, Spain .............................................................. 12
Figure 5  Example of a lateral forcedisplacement cyclic behaviour of a RC Column [37] .. 13
Figure 6  Interaction surface for biaxial load column .......................................................... 14
Figure 7  Experimental tests on RC rectangular columns according to function of the
loading conditions: uniaxial or biaxial; and constant (CAL) or variable axial load
(VAL). Statistics based on a review of the literature and test databases ..................... 17
Figure 8  Column test configurations ................................................................................... 18
Figure 9 Load paths used by different authors .................................................................. 20
vi
Figure 10 Otani et al. Test SP7 [32, 39]: Measured displacements force paths and
hysteresis loops ............................................................................................................. 22
Figure 11 Bousias Test S9 [43]: (a) measured displacement and force path; (b) Hysteresis
loops; and (c) phase lag between measured force and imposed displacement .............. 23
Figure 12 Kawashima et al. [44]: Lateral force vs. lateral displacement hysteresis under
cyclic loadings ............................................................................................................... 24
Figure 13 Oliva and Clough [70]: Test frame and column section dimension ................... 27
Figure 14 Oliva and Clough [70]: Moment vs base rotation results from a rectangular
column .......................................................................................................................... 28
Figure 15 Element model approaches for nonlinear numerical modelling of RC
beam\column ................................................................................................................ 30
Figure 16 Fibre element: distribution of control sections and section subdivision into
fibres (adapted from [104]) ........................................................................................... 32
Figure 17 Triaxial spring model [18] .................................................................................. 33
Figure 18  General scheme of the column specimens and testing directions ....................... 38
Figure 19  RC column specimens dimensions and reinforcement detailing: a) crosssections
details; b) specimen dimensions and general scheme of the reinforcement layout ....... 40
Figure 20 Construction of the columns: a) formwork and column footing; b) column
connection; c) stirrups in place; d) column casting; e) concrete vibration with a
vibrating needle; f) specimens after removing the framework ...................................... 41
Figure 21 Steel stressstrain curves: a) columns N01N04; b) columns N05N16 c) columns
N17N24 ........................................................................................................................ 44
Figure 22 Type displacement paths ................................................................................... 46
Figure 23  Testing setup at LESE laboratory: schematic layout ........................................ 47
Figure 24 Testing setup at LESE laboratory: general view ................................................ 47
Figure 25 Axial load actuator and sliding steel plates at the top column ......................... 48
Figure 26 PXI controller and data acquisition systems ..................................................... 49
Figure 27  Displacement instrumentation scheme adopted: a) lateral displacement
transducers for the EastWest direction; b) lateral displacement transducers for the
NorthSouth direction lateral displacement and relative displacement transducers
along the East and West face ....................................................................................... 50
vii
Figure 28 External steel frame to support the LVDTs to measure the horizontal
displacements in the two orthogonal directions ............................................................ 51
Figure 29 Inverted rollers to support vertical transducers in the West face of the column
to measure the relative displacements along the column height in biaxial tests: a)
Generic view; b) detail view .......................................................................................... 51
Figure 30 Example of the crack in the base of the column (specimen PB01N09) ............. 55
Figure 31 Crack evolution in the East face of column PB01N09 ...................................... 55
Figure 32 Damage states: a) cracking; b) spalling at the column corner; c) spalling at the
overall column section width and longitudinal reinforcement buckling; d) corner bars
rupture .......................................................................................................................... 56
Figure 33 Stirrup failure in column PB12N11 due to hoop rupture ................................. 56
Figure 34 Damage evolution in the columns N15 and N24, subjected to different axial
load ratios but to the same horizontal loading path ..................................................... 57
Figure 35 Plastic hinge length (adapted from [36]) ............................................................ 58
Figure 36 Examples of damage distribution in a 30 x 50cm RC column subjected to
different uniaxial and biaxial loading paths .................................................................. 58
Figure 37 Drift associated with each damage state for different load paths (uniaxial and
biaxial) .......................................................................................................................... 61
Figure 38 Correlation between the ratios (biaxial/uniaxial) of drift demand corresponding
to concrete spalling and bar buckling ........................................................................... 61
Figure 39 Influence of the axial load on the drift associated with each damage state for
biaxially loaded columns ............................................................................................... 62
Figure 40 Influence of the cycle repetition on the drift associated with each damage state
for biaxially loaded columns .......................................................................................... 62
Figure 41 experimental drifts for each damage state and drift limits according to the
FEMA 356 [90], VISION 2000 [130] .............................................................................. 64
Figure 42 Global results of rectangular column PB12N07 for rhombus load path ........... 65
Figure 43 Global results of rectangular column PB12N12 for rectangular load path ....... 65
Figure 44 Global results of rectangular column PB18N12 for circular load path ............. 66
Figure 45 Envelopes for different load paths ...................................................................... 68
Figure 46 Envelopes for different levels of axial force ........................................................ 69
Figure 47 Envelopes for same load path with and without cyclic repetition ..................... 69
viii
Figure 48 Effect of the biaxial load path in the maximum columns strength .................. 69
Figure 49 Proposed method for the definition of yielding displacement ............................ 74
Figure 50 Phase lag between measured horizontal forces and imposed horizontal
displacements for circular load paths ........................................................................... 76
Figure 51 Correlation between maximum and yielding strength ....................................... 77
Figure 52 Strength degradation under cyclic loading ........................................................ 78
Figure 53 Normalized strength degradation (2
nd
and 3
rd
cycle) for different load paths ... 79
Figure 54 Normalized strength degradation (2
nd
and 3
rd
) cycle for different levels of axial
load ............................................................................................................................... 80
Figure 55 Stiffness degradation .......................................................................................... 81
Figure 56 Stiffness degradation for different load paths .................................................... 82
Figure 57 Stiffness degradation for different levels of axial load ....................................... 82
Figure 58 Comparison of cumulative dissipated energy for columns with different load
paths (uniaxial and biaxial loads) ................................................................................ 85
Figure 59 Individual cycle energy and cumulative dissipated energy for rectangular
columns (N09 to N12 and N18) .................................................................................... 86
Figure 60 Individual cycle energy and cumulative dissipated energy for square columns
(N13 to N16) ................................................................................................................. 86
Figure 61 Evaluation of total energy dissipated of columns tested for uniaxial and biaxial
loads (with different load paths) .................................................................................. 88
Figure 62 Evaluation of total energy dissipated for columns with different axial loads .... 89
Figure 63 Evaluation of total energy dissipated for columns with and without cycle
repetition....................................................................................................................... 89
Figure 64 Normalised dissipated energy vs displacement ductility .................................... 90
Figure 65 Damping for a hysteretic halfcycle ................................................................... 92
Figure 66 Equivalent damping vs. maximum ductility demand and bestfit logarithmic
curves from column N13 ............................................................................................... 92
Figure 67 Bestfit equivalent damping vs. maximum ductility demand for different load
paths ............................................................................................................................. 94
Figure 68 Bestfit equivalent damping vs. maximum ductility demand for different levels
of axial force ................................................................................................................. 95
ix
Figure 69 Bestfit equivalent damping vs. maximum ductility demand for the same load
path with and without cyclic repetition ........................................................................ 95
Figure 70 R
2
results for the equation proposals for equivalent damping and each uniaxial
test result ...................................................................................................................... 98
Figure 71 R
2
results for the equation proposals for equivalent damping and all uniaxial
test result ...................................................................................................................... 99
Figure 72 Equivalent damping estimated with empirical expressions and results for all
uniaxial tests ................................................................................................................. 99
Figure 73 Equivalent damping for biaxial tests ............................................................... 101
Figure 74 Best fitted proposals for biaxial equivalent damping ....................................... 101
Figure 75 Ultimate drift capacity: Experimental vs Analytical (EC83 expressions) ...... 104
Figure 76 Damage index evolution for the uniaxial tests ................................................. 108
Figure 77 Contribution of the maximum deformation component to the DI................... 108
Figure 78 Curves for evaluation of resultant displacement .............................................. 110
Figure 79 Damage index evolution calculated for each biaxial test with Equation (27) . 113
Figure 80 Damage index evolution for the biaxial tests according with Equation (33) ... 113
Figure 81 Fibre based modelling (adapted from [174]) .................................................... 118
Figure 82 Modelling strategies with corresponding control section: a) lumpedplasticity
element; b) distributed inelasticity element with forcebased formulation; c)
distributed inelasticity element with displacementbased formulation ....................... 119
Figure 83 Sheardrift envelopes for 30x50cm
2
columns (measured and calculated) ......... 122
Figure 84 Correlation coefficients (R
2
) for the sheardrift envelopes between experimental
and numerical results: a) uniaxial texts; b) biaxial tests strong direction (X); c)
biaxial tests weak direction (Y) ............................................................................... 123
Figure 85 Initial stiffness ratio between experimental and numerical results: a) uniaxial
texts; b) biaxial tests strong direction (X); c) biaxial tests weak direction (Y) ... 124
Figure 86 Secant stiffness evolution for columns PB01N1, PB02N6, PB12N7 and PB12
N12: experimental and numerical results .................................................................... 125
Figure 87 Tangent stiffness evolution for columns PB01N1, PB02N6, PB12N7 and
PB12N12: experimental and numerical results .......................................................... 126
Figure 88 Sheardrift response for columns PB01N1, PB02N6, PB12N7 and PB12N12:
experimental and numerical results ............................................................................. 128
x
Figure 89 FDE index obtained from the comparison of numerical and experimental shear
force evolutions: a) uniaxial tests; b) biaxial tests strong direction (X); c) biaxial
tests weak direction (Y) .......................................................................................... 130
Figure 90 Evolution of the cumulative dissipated energy for columns PB01N1, PB02N6,
PB12N7 and PB12N12: Experimental and numerical results .................................. 131
Figure 91 Ratio between the total dissipated energy obtained with the numerical models
and with the experimental value: a) uniaxial tests; b) biaxial tests total (X + Y); c)
biaxial tests strong direction (X); d) biaxial tests weak direction (Y) ................. 133
Figure 92 Pinching effect in the CostaCosta [179] model ............................................... 137
Figure 93 Unloading stiffness in the CostaCosta [179] model ........................................ 137
Figure 94 Strength degradation in the CostaCosta [179] model ..................................... 138
Figure 95 Inverse problem scheme ................................................................................... 143
Figure 96 Column crosssection and definition of the lateral loading direction .............. 144
Figure 97 Convergence evolution of the cascade optimization strategies in the parameter
identification ............................................................................................................... 147
Figure 98 Relative Global Error of the simplified model results, with (filled marks) and
without (unfilled marks) interaction function, compared with the refined numerical
results.......................................................................................................................... 147
Figure 99 Examples of pushover curves for different columns and different pushover
loading angles for the refined numerical model, the simplified model without the
biaxial bending interaction function and the simplified model with the interaction
function ....................................................................................................................... 148
Figure 100 Baseshear versus drift of columns N05 and N06 Uniaxial tests ................ 150
Figure 101 Baseshear versus drift of column N07 Biaxial test, rhombus displacement
pattern ........................................................................................................................ 151
Figure 102 Baseshear versus drift of column N04  Biaxial test, rhombus displacement
pattern ........................................................................................................................ 151
Figure 103 Baseshear versus drift of columns N09 and N10 Uniaxial tests ................ 152
Figure 104 Baseshear versus drift of column N11 Biaxial test, rhombus displacement
pattern ........................................................................................................................ 152
Figure 105 Baseshear versus drift of column N12 Biaxial test, quadrangular
displacement pattern .................................................................................................. 153
xi
Figure 106 Baseshear versus drift of column N13 Uniaxial test .................................. 153
Figure 107 Baseshear versus drift of column N14 Biaxial test, rhombus displacement
pattern ......................................................................................................................... 153
Figure 108 Baseshear versus drift of column N15 Biaxial test, quadrangular
displacement pattern ................................................................................................... 154
Figure 109 Baseshear versus drift of column N16 Biaxial test, circular displacement
pattern ......................................................................................................................... 154
List of Tables
Table 1 Geometric characteristics of rectangular RC columns tested under biaxial loading
(adapted and updated from CEB 220 [32]) ................................................................... 19
Table 2 Nominal values to the concrete considered in the design...................................... 42
Table 3 Results from compression tests on concrete specimens ......................................... 42
Table 4 Nominal mechanical properties of steel S400 [121] ............................................... 43
Table 5 Results from the tensile tests on steel samples ..................................................... 44
Table 6 Test series identification, specimens main data and displacement paths ............ 46
Table 7 Length of the damage zone ................................................................................... 59
Table 8 Structural Performance Levels and Damage for RC columns adapted from FEMA
356 [90] .......................................................................................................................... 63
Table 9 Drift limits according to the (a) FEMA 356 [90], (b) VISION 2000 [130] ............ 63
Table 10 Drift range obtained from experimental tests ..................................................... 64
Table 11 Summary of test results for columns N01 to N04 ............................................... 70
Table 12 Summary of test results for columns N05 to N08 and N17 ................................. 70
Table 13 Summary of test results for columns N09 to N12 and N18 ................................. 71
Table 14 Summary of test results for columns N13 to N16 ............................................... 71
Table 15 Summary of test results for columns N19 to N24 ............................................... 72
Table 16 Best fit logarithmic curve for global damping for tested each column ............... 96
Table 17 Ultimate drift ratios (experimental/analytical values according to EC83) ...... 104
Table 18 Calculated damage index vs. observed damage ................................................. 107
Table 19 Estimated parameters for RC column damage index calculations .................... 107
Table 20 Damage index ranges for each damage state adopted in the scoring procedure111
xiv
Table 21 Estimated parameters for RC column damage index calculations ................... 111
Table 22 Scores obtained for each column with the different biaxial DI equations ........ 112
Table 23 Concrete mechanical parameters for the numerical model ............................... 120
Table 24 Steel mechanical parameters for the numerical model ...................................... 121
Table 25 Crosssection properties .................................................................................... 145
Table 26 Parameters achieved by the cascade optimization strategy ............................. 146
Chapter 1
Introduction
1.1 General
Looking back over the last decades, it is clear that natural disasters are mostly
associated with earthquakes, volcanic eruptions, floods, tsunamis and cyclonic
storms and that, most of the time, one natural disaster can induce another, for
example tsunamis and earthquakes, or cyclonic storms and floods. These disasters
affect many human lives, and cause high financial loss.
In the particular case of earthquakes, in the last year several examples have
supported the need for investigation to improve the safety of existing constructions,
and have become an important source of information that can help on improving
new design methodologies in the evolution of codes and guidelines. For this reason,
engineers, architects, planners, builders, contractors and authorities all have an
interest in learning from the past. The interest of the scientific community is
revealed in the number of papers concerning lessons learned from the study of
earthquakes published in International Conferences and journals. These include, for
example, the 14
th
World Conference on Earthquake Engineering which included a
topic about Lessons Learned from Recent Earthquakes, and at which 73 papers
were presented, including 41 at the Special Session about the Wenchuan
Earthquake.
Based on observations of building behaviour during earthquakes, Patel [1] points
out twelve lessons for structural engineers: i) codes of practice and ethics must be
strictly followed, and threedimensional analysis should be used in a way which
considers all types of loads and their worstcase combinations; ii) ductile detailing
must be considered; iii) interaction with architects during the early stages of
2 Chapter 1
planning is essential for proper management of structural elements; iv) structural
stability and performance should not be compromised by economy and aesthetics;
v) interaction with professional and academic institutes should be maintained for
recent developments; vi) full documentation, including calculations, drawings, etc.,
must be maintained; vii) fully and detailed drawings should be prepared; viii) there
must be insistence on soil investigation, frequent testing of cement, concrete, steel
reinforcement; ix) reinforcements must be checked against detailed drawings prior
to concreting; x) quality control and site supervision by experienced technical
personnel is very important; xi) there should be an emphasis on weigh batching
and controlled machine mixing of concrete; xii) vibrators should be used at least to
test the foundation, columns and beams.
Bertero [2] performs an identification and analysis of the problems in the USA,
based on the lessons learned from the earthquakes in Chile and Mexico in 1985.
Recognizing that, after the analysis of the US regulations, the performance of
buildings cannot only be enhanced by increasing the seismic loads, the design
should take into account the threedimensional capacity of the soilfoundation
building system and not simply the bare superstructure. Even with improvements
in the design procedures, the greatest need is in the improvement of construction
and maintenance procedures. In fact, many authors point out the need of
improving the quality of construction, in particular the quality of both materials
and workmanship [39].
The earthquake events and the observed damage has resulted in an evolution in the
codes and guidelines; for example, after the Loma Prieta earthquake in 1989,
structurally inadequate behaviour has led the US authorities to revise and propose
new codes for the seismic safety of buildings and bridges [10]. After the Kobe
earthquake in 1995, a study developed by Ohbayashi Corporation [11] has showed
an influence of the seismic codes: in Japan, the first seismic codes date back from
the 1920s and have undergone important revisions in 1971 and recently in 1981.
The study verifies that 6% of the construction after 1981 suffered severe damage,
and 36% of construction prior to 1971 suffered similar damage. For the buildings
with moderate damages, this effect was not so evident: the moderate damage was
identified in 22% of the buildings prior to 1971, and in 11% of the buildings
constructed after 1981.
In contrast, Spence [12] observed that, despite the several earthquakes that have
occurred over many years in Iran and the evolution of the National Seismic Codes,
the country appears powerless in considering earthquake risk, and it would appear
that seismic damage has increased rather than decreased in recent decades.
In Portugal, after the Lisbon earthquake of 1755, and later after the Benavente
earthquake in 1969 [13], seismic safety was taken into account in construction.
However, it was only in 1958 that the first design code concerning the seismic
Introduction 3
safety of buildings [14] was published. This document was used until 1961, with the
introduction of the RSEP [15], and later (1983) replaced by the RSA [16], which
has been used during these years for the structural design. Recently was introduced
the European seismic standard for the Design of Structures for Earthquake
Resistance (Eurocode 8) [17].
The actual seismic design codes highlight linear elastic analysis as a reference for
the design of new structures; however it is essential to determine structural
properties such as initial stiffness, ultimate capacity and different global and local
ductility demands, for evaluating accurately the seismic response [18] which
requires nonlinear dynamic analyses. In order to perform nonlinear dynamic
analysis, researchers have developed various modelling strategies, with different
levels of refinement and complexity. The option for each category depends on the
user objective, and the reliability needed in the representation of the structural
behaviour.
Even with the evolution of these modelling strategies, open problems and questions
regarding the tridimensional nonlinear seismic response and performance of
reinforced concrete (RC) structures still exist [19]. Buildings are tridimensional
structures and the simplification of the tridimensional into bidimensional models
without much loss of accuracy is rather difficult or even impossible. In many
situations, biaxial structural interaction and torsional oscillation may arise as a
result of structural irregularity, thus affecting the structural response [20].
Structural response during recent earthquakes has indicated that the majority of
column failures were caused by high shear stresses, lack of concrete confinement
and bidirectional load effects [21]. It is clear that the earthquakerelated damage to
reinforced concrete elements, due to multiaxial excitation, is much more serious
when compared to uniaxial excitation. The damage caused in one direction affects
the structural seismic performance in the other direction. In fact, even when the
earthquake action is assumed in a single direction, the existent structural
irregularities in many buildings induce a biaxial behaviour.
Therefore is very important to study the threedimensional behaviour of buildings,
and the recent seismic codes and guidelines point the threedimensional dynamic
analysis as the most accurate, that should drive the future of structural
engineering. Nevertheless, many difficulties are still identified in the numerical
representation of the tridimensional response of buildings, motivating further
research. In fact, no efficient simplified hysteretic models are available for the
biaxial representation of the columns' response, and only sophisticated models can
achieve an acceptable representation of columns behaviour [22].
4 Chapter 1
1.2 Research programme objectives
Given the importance of the threedimensional behaviour of reinforced concrete
(RC) structures, in particular the biaxial bending behaviour of reinforced concrete
slender columns, the main objective of the present study was the development of
experimental assessment of RC columns subject to uniaxial and biaxial horizontal
loading, the evaluation of existing models for the simulation of the experimental
tests and the improvement of a simplified hysteretic model to predict the cyclic
behaviour of columns under biaxial cyclic loads.
The experimental programme was designed to study RC columns under flexural
behaviour focusing on the effect of threedimensional loading histories, for different
column geometries. The results of the experimental tests were a major
improvement in the available data concerning the biaxial bending behaviour of RC
elements, and provided valuable information for the development and validation of
the proposed simplified models.
Typical numerical strategies were applied for modelling the experimental tests in
order to verify the models ability to represent accurately the response of columns
to biaxial demands.
Finally, the improvement of simplified global models for the consideration of the
cyclic biaxial bending of RC columns, combined with axial load, is intended to
provide one more step towards the definition of a reliable simplified model to be
used in nonlinear.
1.3 Thesis scope and organization
This thesis is structured into seven chapters. A review of previous research related
to experimental and numerical behaviour of RC columns is provided in Chapter 2.
Chapters 3 and 4 are devoted to the experimental programme performed in the
scope of this Thesis. Chapter 5 deals with the numerical analyses performed with
existent approaches. Chapter 6 describes the development and calibration of the
simplified model to represent the nonlinear biaxial behaviour of RC columns.
Finally, Chapter 7 summarizes the main conclusions of the research work, and
identifies possible future research directions. The Appendixes include further details
on specific topics which are deemed unnecessary to include in the main text for an
adequate understanding of the performed work.
Chapter 2 presents an overview of the most common causes of seismic damage and
failure modes associated with RC columns in recent earthquakes, and reviews the
current state of experimental knowledge of concerning the biaxial behaviour of RC
columns. The available numerical strategies for nonlinear modelling of reinforced
Introduction 5
concrete elements under biaxial bending are briefly reviewed and special attention
is given to member models.
Chapter 3 describes all aspects relating to the test programme of 24 cantilever
columns tested under constant axial load and horizontal biaxial cyclic loading,
including the test specimen details, material properties, test specimen construction,
test setup, loading procedure and instrumentation scheme.
Chapter 4 provides the analysis of experimental results obtained through the tests,
described in Chapter 3, comparing the uniaxial with biaxial response and assessing
the effect of the biaxial load path. The effects of several response parameters on the
column flexural response are evaluated, such as damage evolution, ultimate
displacement, energy dissipation, stiffness and strength degradation, and viscous
damping.
The evaluation of existent numerical models currently used for predicting the
seismic performance of reinforced concrete columns is presented in Chapter 5. To
this end, the experimental results obtained in the experimental programme
described in Chapters 3 and 4 are compared with the results predicted using three
modelling strategies, based on elements with lumpedplasticity, elements with
distributed inelasticity and forcebased formulation, and elements with distributed
inelasticity and displacementbased formulation.
Chapter 6 is devoted to the description of a simplified hysteretic model developed
for the forcedeformation behaviour of reinforced concrete members under biaxial
loading with axial force. It embodies an existing uniaxial hysteretic model with
piecewise linear behaviour as the basis for its development, and proceeds from
analogy and comparison with the biaxial formulation of the BoucWen smooth
hysteretic model. The validity of the model is demonstrated through the analytical
simulation of the biaxial tests presented in Chapters 3 and 4.
Finally, the most relevant results and conclusions of this work are summarized in
Chapter 7. Possible future research directions are also suggested.
Chapter 2
Previous Research and
Background
2.1 Introduction
Since columns are key structural elements for the seismic performance of buildings,
special attention should be given to their structural response under load reversals.
Moreover, earthquake effects generally require the inclusion of two horizontal
component loads that are recognized to be more damaging than single direction
actions. The interest in the inelastic response of axially loaded members under
biaxial bending moment histories is relatively recent, and the available
experimental results are limited. This is possibly due in part to the uncertainty of
combining histories of bending moments in the two orthogonal directions, adding
considerable complications to the problem.
The practical result is that our presentday knowledge of the inelastic behaviour of
RC columns under biaxial cyclic moments is very much behind to our
understanding of their behaviour under uniaxial cyclic bending with axial load. In
fact, besides the fibre based models, the existing simplified analytical models are
not mature enough to be incorporated into code standards, by contrast with
uniaxial simplified global models which are already accepted in international codes.
8
This
work
open
2.1.
The
obser
majo
defici
const
Varu
build
bond
capac
joints
vertic
prope
and f
Some
defici
throu
Figu
the
Chapter ai
ks in the fiel
problems.
.1 Typica
during
earthquake
rvation of p
ority of cases
iencies of
truction def
um [24] poin
dings: i)
d/anchorage
city; iv) in
s; vi) influe
cal and hor
erties; viii)
finally, x) st
e of these c
iencies, for
ugh the cont
ure 1  Flexur
2009 LAqui
ms at provi
ld of biaxia
al cause
g earthq
performanc
past seismic
s, by the fai
RC buildi
ficiencies as
nts out the
lack of
/lapsplices
nadequate fl
ence of the
rizontal irre
higher mod
tructural de
categories a
which the
tribution of
ral deficient b
ila Earthquak
iding a brie
l behaviour
s and co
quakes
ce of RC bu
events. The
ilure of the
ings can b
well as wit
ten most c
stirrups/h
slipping a
lexural cap
infill mason
egularities, a
des effects;
ficiencies du
re in fact r
extensive d
many of th
behaviour in
ke; b) rectan
earthquake
ef review of
of RC colu
onsequen
uildings has
e collapse o
vertical mem
be associat
th deteriora
common ca
hoops, con
and bond
acity; v) in
nry on the
abrupt chan
ix) strong
ue to archite
related to d
damage or
hese causes.
RC columns
ngular column
e in Lorca, S
a)
past experi
umns, and t
nces of c
s been well
of a RC bui
mbers [23].
ed with d
ation and st
auses of fail
nfinement
splitting;
nadequate s
seismic beh
nge in struc
beam weak
ectural requ
design, deta
collapse of
: a) in the Sa
n in residenti
Spain
imental and
o identify t
olumn fa
documented
lding is cau
The seismic
design, deta
tructural mo
ure or dam
and duc
iii) inadequ
shear streng
haviour of f
ctural and/
kcolumn m
uirements.
iling and co
f structures
an Salvatore
al building in
Chapter 2
d numerical
the existing
failure
d from the
used, in the
c behaviour
ailing and
odification.
mage in RC
ctility; ii)
uate shear
gth of the
frames; vii)
/or element
mechanisms;
onstruction
can occur
hospital, in
n the 2011
b)
b)
Previous Research and Background 9
Columns properly designed for earthquake loading are able to prevent brittle
failure and to have a ductile behaviour. In fact adequate reinforcement and
detailing is very important in providing the ductile behaviour for the RC
elements [25]. Moreover, the flexural behaviour is conditioned by the axial force
and by the amount of reinforcement in the plastic hinge region. Figure 1 shows
examples of RC columns with failures associated with inadequate flexural capacity,
although in these examples shear and confinement deficiencies are also evident
(transversal reinforcement with large spacing at the columns extremities).
Deficient flexural behaviour can be more evident in exterior corner columns where
varying levels of axial force can be expected during earthquakes, leading to high
levels of axial force. In some cases it may be difficult to differentiate flexural
compression and shear compression failure, as both take place in or near the
column ends and involve crushing [23].
The inadequate transversal reinforcement in terms of size, spacing and detailing is
the principal cause of the shear failure of RC columns. This problem can be
increased in exterior corner columns in buildings with inplan irregularities [24].
The columns with shear failure commonly exhibit a diagonal fracture in the column
mid zone, as illustrated in Figure 2a) and b). Another typical cause of shear failure
is related with insufficient area of transversal reinforcing steel, with wide spacing
and deficiently anchored to the concrete core, where ties should have enough length
or proper bar bents to promote adequate splice anchorage (see Figure 2c).
Beamcolumn connections play also an important role in the seismic behaviour of
RC buildings; poor behaviour of beamcolumn joints can lead to the collapse or
severe damage of buildings during earthquakes. In some cases, the beam
longitudinal reinforcement is not properly anchored in the beam column joint [23]
(see Figure 3a) or there is no adequate transverse reinforcement in the joints
(see Figure 3b and c).
10
The
not c
[28].
often
short
cause
stairc
devel
can l
relate
groun
c)
Figu
buildin
circular
c) defic
contribution
considered in
In the case
n observed. T
t portion of
ed by openi
cases. If thi
loped with i
lead to she
ed to the ab
nd floor sto
ure 2  Shear
ngs in the 20
column [27];
cient flexural/
building
n of nonstr
n the design
e of the infil
The first is
clear colum
ngs in the i
is effect is n
increased st
ear failure o
bsence of th
orey due to
a)
deficient beh
09 LAquila
failure is rel
/shear behav
g in the 2011
ructural elem
n of new bui
ll masonry
associated w
mn, creating
infill walls,
not consider
iffness whic
of the colum
he infill mas
o its use fo
haviour in RC
Earthquake:
lated to the t
viour in a corn
earthquake i
ments to th
ildings and
panels, two
with cases w
g the socall
for doors or
red in the d
ch becomes s
mn, as illus
sonry panel
or car park
C columns in
a) square co
transversal re
ner column in
in Lorca, Spa
he structura
in the asses
o principal m
where maso
led short co
r windows,
design, a sh
subjected to
strated Figu
ls in one sto
king or com
b)
c)
residential
lumn [26]; b)
einforcing stee
n a residentia
ain
al response i
ssment of ex
mechanisms
nry infill wa
lumn: this s
or for landi
ort column
o shear force
ure 4a. The
orey, freque
mmercial ar
Chapter 2
)
eel;
al
is typically
xisting ones
have been
alls leave a
situation is
ing slabs of
is actually
e level that
e second is
ently in the
reas, which
Previous Res
induces an
storey me
masonry p
fact can in
the outer f
Figu
anch
colum
search and Ba
n abrupt ch
echanism [2
panels can d
nduce additi
frames [31].
ure 3  Inadeq
horage of bea
mn joint in a r
buildi
ckground
ange of the
29, 30]. Mo
drive torsion
ional uncons
quate joint be
m longitudin
residential bu
ng collapse in
storey stiff
oreover the
n phenomen
sidered force
a)
ehaviour in r
nal reinforcem
uilding (2009
n the 2011 ea
fness, leadin
asymmetri
na not predi
es, especiall
residential bu
ment b) defici
earthquake
arthquake in
ng to a pote
ic distribut
icted in the
y in the con
uildings: a) in
ient detailing
in LAquila,
Lorca, Spain
ential global
ion of the
e design, and
ncrete colum
c)
nappropriate
g of a beam
Italy); c) RC
n
11
l soft
infill
d this
mns of
b)
C
12
Figu
caus
2.1.
The
their
impo
force
prope
reinfo
In fa
steel)
consi
other
struc
eleme
relev
varia
exter
load
The
decad
nonl
study
failur
betw
bidir
ure 4  Colum
sed by windo
.2 Hyste
seismic beh
s elements
ortant role
s due horiz
erties, such
orcement, a
act, these s
), are the m
idered as in
r factors c
ctural eleme
ents. The p
ant issue
ations in th
rnal column
level.
seismic beh
des many
linear hyste
y, such as:
re of the c
ween concre
ectional) [32
mn failure due
ow openings i
and
retic beh
haviour of R
and the c
because the
zontal loads
h as dimens
re directly i
section char
most relevan
ndividual me
an influenc
ents (mason
position of
because du
he axial forc
ns, and espe
haviour of R
studies hav
eretic behav
i) the conf
olumns; iii)
ete and ste
236].
e the presence
in residential
b) 2011 eart
haviour
RC building
connections
ey support
s, like wind
sions, amou
influenced b
racteristics
t in the col
embers. In t
ce its beha
ry infill pan
the column
uring earth
ce level of
cially corne
RC columns
ve been pe
viour; severa
finement of
) the const
eel; v) the
e of nonstru
building: a)
thquake in Lo
of RC c
gs is depend
between t
the weight
d or earthqu
unt of longi
by the level
and the m
umn behavi
the case of
aviour, as
nels, stairs)
n in the b
quakes inte
(that can b
er ones, may
s is controll
erformed to
al aspects a
f the concre
tant or vary
e direction
a)
uctural elemen
2009 earthqu
orca, Spain
columns
dent on the
hem. The
t of the str
uakes. In th
itudinal rein
of axial and
material pro
iour under h
a column i
for exampl
and the con
building geo
erior colum
be considere
y be subject
ed by many
o character
are consider
ete core; ii)
ying axial
of the loa
nts. Short co
uake in LAq
e behaviour
columns pl
ructure and
he design, t
nforcement
d shear force
operties (con
horizontal lo
ncluded in
le the atta
nnections to
ometry can
mns show o
ed constant
ted to a va
y aspects, a
rize the RC
red as key
) the effect
load; iv) th
ad (unidire
Chapter 2
olumn effect
quila, Italy;
of each of
lay a very
d the shear
the column
and shear
es.
ncrete and
oads, when
a building,
ached non
o the other
be also a
only small
t); however
arying axial
and in past
C columns
factors for
and shear
he slippage
ectional or
b)
Previous Research and Background 13
The present study is focused on the behaviour of slender RC columns, with the
nonlinear hysteric behaviour governed by flexure, and so it is important to recall
the hysteretic characteristics of these elements. This illustrative analysis is based
on a typical forcedeformation response of a RC column subjected to unidirectional
load and constant axial force, as presented in Figure 5. The following issues can be
highlighted: i) the envelope of the force displacement curve can be seen as an
approach to the monotonic response of the section, where two particular points can
be identified: the concrete cracking point (C) and the yielding point (Y), each of
these points correspond to a reduction of the member stiffness; ii) up to yielding
the unloading stiffness is approximately constant, but the dashed lines US1, US2,
US3 and US4 indicates that, after the yielding point, the unloading stiffness
reduces with increase of the inelastic branch; iii) the point P indicated in Figure 5
point to an effect resulting from the closure of the open cracks, slippage and shear
effects, with all factors combined contributing for the reduction of the reloading
stiffness, classically referred as pinching effect; and iv) in the last cycles after a
large incursion on the inelastic behaviour of the RC column, it can be seen that, for
repeated cycles with the same displacement, the loading stiffness and the strength
degradation at the peak displacement tends to decrease between successive cycles
which is normally associated with concrete degradation and buckling of
reinforcement steel. The presented case shows the most relevant inelastic hysteretic
features of the response of a slender RC column subjected to horizontal uniaxial
load.
Figure 5  Example of a lateral forcedisplacement cyclic behaviour of a RC Column [37]
60 40 20 0 20 40 60
80
60
40
20
0
20
40
60
80
P
US4 US3 US1 US2



0 500 1000 1500 2000 2500
60
40
20
0
20
40
60
T
op D
isplacem
ent (m
m
)
Step
T
o
p
f
o
r
c
e
(
k
N
)
Top displacement (mm)
C
Y
1
14
It is
respo
axial
Load
behav
repea
while
reduc
The
conte
direc
recen
biaxi
proce
in th
struc
integ
surfa
Gene
Figur
gener
envel
comb
M
z
).
reinfo
important
onse, such a
force that c
ding history
viour of RC
ated cycles
e internal cy
ctions [40].
design anal
ext of unia
ctions was
ntly the des
ial bending
edures based
he design
ctures need
gration over
ace [42].
erically, the
re 6, where
ric case 3 is
lope surface
binations of
The surfac
orcement st
to be aw
s the mater
can induce p
y is anothe
C sections [
postyieldin
ycles after la
ysis of a rei
axial mome
often achi
sign of reinf
has been
d on compu
[41, 42]. C
to be able
r the concr
e momenta
e the loadin
s described f
e (for exam
f axial force
e depends o
eel.
Figure 6 
are of the
ial and sect
particular p
er very imp
38, 39]. For
ng induce a
arge inelast
inforced con
ents, and t
ieved throu
forced concr
the subjec
utational me
Computer p
e to compu
rete crosss
axial force
ng cases 1
for biaxial l
mple, a failu
load and m
on the sect
Interaction s
influence
tion propert
performances
portant iss
r uniaxial l
a decrease o
ic may exhi
ncrete colum
the combin
ugh simplif
rete column
ct of much
ethods have
programs fo
ute the inte
section, in
interaction
and 2 refer
loading. The
ure surface)
moments in
ion dimensi
surface for bia
of other fa
ies, PDelta
s.
sue which
loading, as
of loading s
ibit a lower
mn was usua
nation of th
fied proced
n under to a
h research
recently be
or designing
ernal forces
order to o
of a colum
r to the un
e interaction
) which is
the princip
ions and am
axial load col
actors on t
a effects, or t
affects the
mentioned
stiffness and
strength an
ally underta
he moment
ures. Howe
axial compr
[41]. Some
een develope
g reinforced
through th
obtain the
mn is summ
niaxial load
n is represen
obtained fo
pal direction
mount and
lumn
Chapter 2
he column
the level of
hysteretic
before, the
d strength,
nd stiffness
aken in the
ts in both
ever, more
ression and
automatic
ed to assist
d concrete
he stresses
interaction
marized in
ding, and a
nted by an
or different
ns (M
y
and
location of
Previous Research and Background 15
For the biaxial loading case, the coupling between the two transverse directions is
very important for the response and also dependent on the load path as observed in
different studies [22, 34, 4347]. Bousias et al. [48] have reported the biaxial effect
in a series of experiments describing the differences between the uniaxial and
biaxial behaviour, especially the reduction of the strength and stiffness with biaxial
bending, and the increase of hysteretic energy dissipation.
Another important issue is the column shear span ratio, for low values (say below
about 2.5 to 4), the column behaviour can be controlled by the inelastic shear
effects [38, 49]. This behaviour leads to a reduction of the energy dissipation
capacity and to a stiffness and strength decrease in the hysteretic cycles [33, 50].
2.2 Experimental studies on the seismic behaviour
of RC columns
2.2.1 Introduction
The experimental study of RC elements under earthquake loads is very important
since it permits the observation and measurement of the particular performance of
the element. Regarding the behaviour of RC elements under lateral loads, a large
number of various types of studies have been carried out in past years which, apart
from the specific subject of each study, have generally covered elements with or
without axial loads, as well as uniaxial and biaxial bending. In the case of zero
axial force, these studies are typically associated with beams and, therefore, most of
them focus only on the uniaxial behaviour of these elements. Columns are studied
with nonzero axial force under uniaxial or biaxial bending and also tested with
constant or variable axial load. In the case of slender elements, the behaviour is
governed by flexure but, depending on the crosssection geometry and the already
mentioned shear span ratio, the shear force can also represent an important subject
for the study. Studies of RC elements can therefore also be divided into elements
where the global behaviour is governed essentially by flexure or by shear, or by
both mechanisms.
In the view of the present study, it is important to perform a retrospective review
of experimental tests performed on rectangular columns under horizontal loads,
especially with biaxial behaviour.
The response of RC members subjected to axial loads together with biaxial bending
moment reversals is recognized as a very important research topic for building
structures in earthquakeprone regions. On one hand, studies of the response of RC
building columns to earthquake actions deal in general with its threedimensional
response, due to the random characteristics of the earthquake direction, and to the
16 Chapter 2
actual building irregularities. On the other hand, the biaxial features of bending
moment histories applied to a given RC column section tend to reduce its actual
capacity and to accelerate the strength and stiffness deterioration process during
successive load reversals. In addition, the 3D response of frame structures to actual
earthquake motions generally does not induce the same type of increased
deterioration in beams because they behave essentially in only one direction
(vertical), i.e. the potential formation plastic hinges in beams are not aggravated
by that fact. This means that both the biaxial loading effects in columns and the
3D features of the general structure response both positively contribute to
inelasticity and damage concentration in the columns rather than in the beams,
which is essentially the opposite of presentday design code requirements created to
avoid collapse of RC frame structures under lateral load reversals (plastic hinges in
the beams rather than in the columns) [32].
Experimental research work on the inelastic response of RC members under
compression axial force and biaxial lateral cyclic bending loading conditions is
currently very limited. Uncertainties concerning the relationship and combination
of the two orthogonal horizontal loading paths, associated to the complexity of the
experimental setup, certainly justified this lacuna.
As a consequence, current knowledge concerning the inelastic response of RC
columns under biaxial cyclic moments is very much less than that of the uniaxial
cyclic bending behaviour with compressive axial load [32, 51, 52].
The available test results for biaxial bending under constant axial load are not so
extensive when compared to those on uniaxial bending, although they have been
delivered over a period of almost 30 years. On the basis of an extensive analysis of
international experimental databases and from a literature review of studies
covering tests on rectangular RC columns subjected to cyclic loading, a statistical
analysis was performed based on the data collected for 453 column tests (see Figure
7). From the data analysis, it was verified that: i) only 27 of these columns (6%)
were tested with variable axial load; and ii) only 56 (12%) were tested in biaxial
loading conditions. These statistical figures emphasize the insufficient experimental
results for the comprehensive characterization of the biaxial cyclic behaviour of RC
columns.
Due to testing difficulties and because there are still open questions regarding the
cyclic behaviour both in biaxial bending with constant axial force and in uniaxial
bending with simultaneously varying axial load, very few experimental studies
have, as yet, tackled the more general problem of biaxial bending with varying
axial force [32, 33], as can be seen in Figure 7. From the literature review, only the
tests performed by Low and Moehle (two columns) [47] and by Chang (two
columns) [53] refer to rectangular crosssections with different dimensions and
Previous Research and Background 17
amount of longitudinal reinforcement in the two principal directions, leading to
different characteristics in terms of stiffness and strength.
In recent years, a very small number of tests have also been performed in beam
columns subassemblages where the columns were with biaxial load [5457].
Figure 7  Experimental tests on RC rectangular columns according to function of the
loading conditions: uniaxial or biaxial; and constant (CAL) or variable axial load (VAL).
Statistics based on a review of the literature and test databases
The next sections presents a summary of the previous works involving experimental
studies performed with biaxial bending. The analyses focus on the main findings
obtained from tests on columns subjected to biaxial lateral loading (with constant
and varying axial force), and under different test conditions: biaxial tests
performed on global bare frame structures are also analysed. All relevant
experimental work performed before 1996 is very well summarized in the CEB
Report N220, RC Frames under Earthquake Loading State of the Art Report
[32], and so the present section will focus on the main findings of the experimental
studies and work performed after 1996.
C
A
L
V
A
L
0
5
0
1
0
0
1
5
0
2
0
0
2
5
0
3
0
0
3
5
0
4
0
0
B
i
a
x
i
a
l
U
n
i
a
x
i
a
l
19
13
54
378
18
2.2.
From
speci
again
the o
confi
and t
The
assum
many
know
The
at m
true
in th
Rega
have
of th
squar
can
perfo
to an
.2 Specim
m the cyclic
imens were
nst rotation
other studie
gurations c
the hammer
double cur
med as the b
y factors, m
wn studies o
cantilevert
midheight of
in columns
he socalled p
arding the c
been tested
he columns a
re crosssect
be used du
orm two uni
nalyse the d
men geom
c biaxial te
used. The
on both en
es report th
an be used,
rhead config
rvature spe
best models
most of the
n the differe
ype model a
f the column
governed b
plastic hinge
Fig
olumn heigh
d as describ
are summari
tion. Althou
ue to the d
iaxial tests
ifferences in
metries
ests describ
prototype t
ds (Figure 8
he use of t
, namely th
gurations [59
cimens wit
s to simulate
authors use
ences betwe
assumes tha
n, and take
by flexural b
e zones loca
gure 8  Colum
ht and cross
bed in Table
ized. As pre
ugh no clear
difficulties i
(one in eac
n the stiffnes
bed in the
test conside
8a) was use
the cantilev
he doubleen
9].
h rigid zon
e a typical b
e the cantile
een these spe
at the inflec
es no damag
behaviour w
ated at the c
mn test confi
ssection dim
e 1 where t
eviously men
r justificatio
in developin
ch principal
ss and stren
literature,
ering the co
d only in on
ver type (F
nded cantile
nes on top
building col
evertype sp
ecimen conf
ction point
ge, which ca
where the da
column ends
igurations
mensions, m
he main geo
ntioned, the
on for a squ
ng the test
direction),
ngth in each
two types
olumn specim
ne study [58
Figure 8b).
ever with fle
and botto
umn. Howe
pecimen. Th
figuration ty
of a column
an be consid
amage is co
s.
many differe
ometric cha
e most colum
uare section
t setup, th
and the ne
direction.
Chapter 2
of column
mens fixed
8], while all
Other test
exibleBase
om can be
ever, due to
here are no
ypes.
n is located
dered to be
oncentrated
ent sections
aracteristics
mns have a
n; mainly it
he need to
ew variable
Previous Research and Background 19
Table 1 Geometric characteristics of rectangular RC columns tested under biaxial loading
(adapted and updated from CEB 220 [32])
Reference
Number of
Specimens
l
[mm]
b x h
[mm]
Axial
load
(C or V)
Takizawa and Aoyama (1976) [60] 4 600 200x200 C
Otani et al. (1980) [39] 4 1372 305x305 C
Takiguchi et al. (1980) [58] 5 250 150x150 C
Park et al. (1984) [61] 1 1600 400x400 C
Low and Moehle (1987) [47]
2 546 127x165 C
2 546 127x165 V
Li et al. (1988) [62]
1 600 200x200 C
4 600 200x200 V
Saatcioglou and Ozcebe (1989) [63]
2 1000 350x350 C
1 1000 350x350 V
Zahn et al. (1989) [64] 2 1600 400x400 C
Bousias et al. (1992) [43]
9 1500 250x250 C
1 1500 250x250 V
Kim and Lee (2000) [65]
2 1200 100x100 C
4 1200 200x100 C
Qiu et al. (2002) [22] 6 900 200x200 C
Tsuno and Park (2004) [66] 2 2750 550x550 C
Bechtoula et al. (2005) [67]
2 625 250x250 C
1 1200 560x560 C
2 1200 560x560 V
2 625 242x242 V
1 1200 600x600 V
Kawashima et al. (2006) [44] 6 1750 400x400 C
Chang (2010) [68] 1 355 750x600 C
C Constant axial load; V Variable axial load
2.2.3 Displacement patterns
The behaviour of RC columns under cyclic behaviour is especially influenced by the
geometric and mechanical characteristics of the column crosssections and the of
axial force level. However, the displacement pattern can also influence the global
behaviour and in terms of biaxial bending forces, the displacement history and the
way the biaxial forces are combined in two orthogonal directions, is a very
important subject. This section will therefore review the displacement patterns
used in published studies, and also briefly analyse their influence on the column
behaviour.
From the literature review it was possible to describe several types of laws
considered in displacement control biaxial tests. Displacement patterns can be
applied only along the X and Y axes, illustrated in Figure 9 as L1 and L2, creating
20
a cru
have
rhom
the f
the p
The
of th
biaxi
differ
study
uniax
patte
In th
displ
zero
the c
betw
Addi
and
inclu
consi
inten
displ
repet
degra
uciform disp
been used
mbus or diam
form of an 8
presented di
use of these
he performed
ial load pat
rent biaxial
y of Saatcio
xial and dia
ern, an earl
he studies
acement pa
displacemen
column resp
ween the two
itionally, th
calibration
uding the b
idered as r
ntion of ca
acement pa
tition of eac
adation alon
placement p
d by differe
mond (L4),
8 (L6) and t
splacement
Figure 9
e (or other)
d biaxial tes
ths, and on
displaceme
oglou et al.
agonal load
ier progress
of Li et a
ath with the
nt was obse
ponse and c
o orthogonal
e experimen
of mathem
biaxial inter
epresentativ
librating n
aths with sm
ch displacem
ng the tests.
L1
L5
pattern. Ot
nt authors,
expanding
the circular
patterns ca
9 Load path
) displaceme
stes: on one
n the other
nt patterns
[63], simil
paths; howe
sive degrada
al. [62] and
e 8 shape, a
erved. The
can induce p
l directions.
ntal results
matical mod
raction. Th
ve of an ea
umerical m
mooth increa
ment level, i
.
ther configu
, namely: t
square (L5)
or elliptical
an be found
hs used by di
ent patterns
e hand, the
r hand the
in the colu
lar hysteret
ever, for th
ation of stre
d of Kawas
a characteris
biaxial load
particular e
are intende
dels that r
he presented
arthquake
models impo
ase of the p
in order to c
L2
L6
urations of
the diagona
), the squar
l (L7 and L
in the litera
ifferent autho
s is based o
comparison
evaluation
mn respons
ic behaviou
e elliptically
ength and s
shima et a
stic drop of
ding history
effects as a
ed to be use
represent th
d displacem
displacemen
oses the de
peak displac
capture the
L3
L7
displacemen
al cruciform
re in each q
L8). Some va
ature.
ors
on the main
between un
n of the eff
e. For exam
ur was foun
y shaped dis
stiffness was
l. [44], bas
the restorin
had a clea
result of th
ed in the de
he column
ment paths
nt path; ho
efinition of
cements, an
strength an
3
7
Chapter 2
nt patterns
m (L3), the
quadrant in
ariations of
n objectives
niaxial and
fect of the
mple, in the
nd between
splacement
s observed.
sed on the
ng force at
ar effect on
he coupling
evelopment
behaviour,
cannot be
owever the
f simplified
nd with the
nd stiffness
L4
L8
Previous Research and Background 21
2.2.4 Behaviour of RC columns under biaxial bending
with constant axial load
As stated before, the available test results for biaxial bending under constant axial
load are not so extensive when compared to those for uniaxial bending, although
they have been delivered over a period of almost 30 years. Contributions are
summarized in Table 1.
In general, most research findings agree that, additionally to the expected
significant influence of axial loads on the hysteretic response of columns, the biaxial
transversal load cycles are responsible for increasing the strength and stiffness
degradation when compared to the uniaxial response. In addition, the failure
mechanism of RC columns is found to be very dependent on the loading
path/history, and strongly affects both the ductile and energy dissipation capacity
of the columns. On the other hand, there is some experimental evidence that
plastic hinge zone lengths tend to be stable at around theoretical values and are
not strongly affected by biaxial loading.
The CEB Report N220, RC Frames under Earthquake Loading State of the Art
Report [32] includes the major findings of the experimental studies about biaxial
flexure with constant axial load of RC columns until 1992. The major findings are
resumed next.
Observing the force path measured with the square displacement path, several
authors [39, 60] have reported the expected elastoplastic behaviour of the RC
columns, for which after yielding, the square force paths show a tendency to cluster
into a single square (see Figure 10).
Several authors [39, 43, 58, 60] reported a rotation of the measured force paths by
about 1020
o
with respect to the applied displacement path (see Figure 10). This
rotation occurs when the displacement in one direction is changed while the
displacement in the opposite direction is maintained at a practically constant level;
the force required in this lateral direction consequently drops. The effect of this
coupling on the hysteretic curves induces an almost vertical unloading branch,
which increases the level of dissipated energy in each loop [32].
22
F
Saatc
with
even
and t
repor
displ
respo
two
result
but i
appe
Low
recta
the b
force
assoc
repor
Figure 10 O
cioglou and
increasing
more round
the imposed
rted for th
acement cu
onse, thus i
directions.
ts, concludi
increases w
aring (see F
w and Moeh
angular cros
beginning of
drop unde
ciated with
rted by Li e
Otani et al. T
d Ozcebe [63
ellipse amp
ded shape ch
d displaceme
he square l
rves means
ncreasing th
For the circ
ing that the
ith the defl
Figure 11).
hle [47] per
s sections. I
f the unload
er constant
the type of
et al. [46].
est SP7 [32,
hyste
3] reported
litude, for w
haracteristic
ent vector. T
load paths.
they are m
he energy d
cular load p
e phase lag
lection amp
rformed the
In the third
ding branch,
zero displa
displaceme
39]: Measure
eresis loops
on the use
which the fo
c of the pha
This fact is
The incre
much wider w
dissipation
path, Bousi
is almost c
plitude and
e first biax
test, beside
, as also rep
acement wa
ent path sim
ed displacem
e of elliptic
orcedisplace
ase lag betw
associated w
ease in rou
when compa
due to the
ias et al. [43
constant at
at the test
xial tests o
es the abrup
ported by ot
as observed
milar to test
ments force pa
al displacem
ement curve
een the resu
with the effe
undness of
ared with th
coupling be
3], have fou
each deflec
t end, when
on two colu
pt drop of t
ther authors
d, this phen
L6. this fac
Chapter 2
aths and
ment paths
es show an
ultant force
fect already
the force
he uniaxial
etween the
und similar
ction level,
n failure is
umns with
the force at
s, a similar
nomenon is
ct was also
Previous Res
Bousias et
not dupli
displaceme
between t
transverse
each of the
each direc
adverse ef
beneficial
dissipation
Kawashim
performan
crosssectio
columns w
elliptic dis
of the Kob
and unloa
bilateral l
directions.
feature sim
search and Ba
F
displa
(c)
t al. [43] foc
icate those
entcontrolle
the three l
directions
e two transv
ction, but a
ffect on the
to the stru
n.
ma et al. [4
nce of reinfor
on, using cy
were loaded
splacement
be and Nort
ding hyster
loading, wh
Only the h
milar to th
ckground
Figure 11 B
acement and
phase lag be
cused on the
of earlier
ed paths.
loading dir
produced a
verse direct
lso increase
e structural
uctural resp
44] studied
rced concret
yclic and hy
unilaterally
paths. For
thridge eart
esis curves
hich results
hysteresis a
at under th
Bousias Test S
d force path; (
etween measu
displacem
e effect of lo
r researche
The study
rections. Th
an apparent
ions when c
ed the hyste
seismic res
ponse, such
the effect
te bridge co
ybrid loadin
y or bilatera
the hybrid
thquakes we
are round
from the
at the first
he unilatera
S9 [43]: (a) m
(b) Hysteresi
ured force an
ment
oad path. T
ers, but co
emphasise
he strong
t reduction
compared w
eretic energ
sponse, som
h as the in
of bilateral
olumns with
ng tests. In
ally using di
loading cas
ere used. In
near the pe
interaction
excursions,
al loading.
measured
is loops; and
nd imposed
The examine
overed diffe
ed the sign
coupling be
in strength
ith the load
y dissipatio
me effects ca
ncrease of h
l excitation
h 1.35m tall
the cyclic
agonal, squa
ses, the grou
the cyclic
eak displace
of restorin
in each loa
This is bec
ed load path
ferent trans
nificant cou
etween the
h and stiffne
ding separat
on. Therefor
an be consi
hysteretic e
n on the se
and 400x40
loading tes
are, circular
und accelera
tests the lo
ements unde
ng forces in
ading step,
cause, unde
23
hs did
sverse
upling
e two
ess in
ely in
re the
idered
energy
eismic
00mm
t, the
r, and
ations
oading
er the
n two
has a
er the
24
circu
first
the c
bilate
respe
after
bilate
capac
deter
in bo
loadi
loadi
devel
has t
Figu
Qiu
biaxi
unde
energ
hyste
the e
that
great
plast
Tsun
2.75m
concl
some
displ
the
Tsun
ularpath loa
excursions a
column und
eral cyclic l
ectively. Fu
3% drift, w
eral hybrid
city of rein
riorate unde
oth cyclic a
ing test, for
ing excursio
loped in the
to be careful
ure 12 Kaw
et al. [22] t
ial loading w
er biaxial lo
gy dissipatio
eresis dissipa
energy para
under unid
ter than tha
ic deformat
no and Park
m tall and
lude that th
e cyclic load
acement of
same colum
no and Park
ading, the c
at each load
der bilatera
loading is 1
rthermore,
while deteri
d loading (
forced conc
er the bilate
and hybrid
which the
ons at eac
e hybrid loa
lly determin
washima et al
tested seven
with differe
ads, were fo
on capacity.
ation energy
ameter of a
directional l
at under un
ion capacity
k [66] also p
with 550x5
he plastic h
dings and i
a column i
mn subject
k, in agreem
column was
ding step (s
al hybrid lo
1423% and
deterioratio
ioration did
not showed
rete bridge
eral excitati
loading tes
loading disp
ch loading
ading test. I
ned in the cy
l. [44]: Latera
cycl
n specimens
ent load pat
ound to we
According
y relates clo
specimen u
oading. The
niaxial load
y.
performed c
550mm cros
hinge zone l
is not affec
in a bidirec
ted to the
ment with t
s first loade
see Figure 1
oading, the
d 1535% sm
on of the lat
d not occur
d herein).
columns w
ion in comp
sts. Failure
placement w
step, is m
In fact, load
yclic loading
al force vs. la
lic loadings
s of reinforc
ths. The int
aken the bi
to the test
osely to the
under biaxia
e specimen
ding, which
cyclic bidire
ss section.
ength tends
cted by bid
ctional cycli
e standard
the Ohno a
ed only in t
12). Compar
e strength
maller in N
teral restori
until 6.3%
Flexural st
with a squar
parison with
of the colu
was increase
more extens
ding protoco
g tests.
ateral displace
ced concrete
teractions o
iaxial streng
result, altho
loading pos
al load is a
damage un
agrees with
ctional test
In this stu
s towards t
directional l
ic loading is
unidirectio
and Nishiok
the N direct
ring to the
of the colu
S and EW
ing force is
peak drift
trength and
re section si
h unilateral
umns under
ed stepwise
sive than t
ol is import
ement hyster
e column su
of biaxial de
gth and the
ough the ac
sition and p
apparently l
nder biaxial
h the reduct
s two on R
dy it was p
heoretical v
loading. Th
s smaller th
onal loadin
ka [69], stat
Chapter 2
tion in the
strength of
umn under
directions,
significant
under the
d ductility
ignificantly
l excitation
r the cyclic
with three
the failure
ant, and it
resis under
ubjected to
eformation,
e hysteretic
ccumulative
path length,
larger than
l loading is
tion of the
RC columns
possible to
values after
he ultimate
han that of
ng pattern.
te that the
Previous Research and Background 25
total dissipated energy until the ultimate state is achieved are approximately the
same for biaxial and uniaxial conditions, meaning that the biaxial loading did not
affect the total energy dissipation capacity of a column that reaches failure.
Chang [68] pseudodynamically tested two reinforced concrete bridge columns at
2/5 reduced scale, and one identical column was also tested under biaxial cyclic
horizontal load with constant axial load. The results showed that the distinct
characteristics of biaxial hysteretic loops are rounded at corners and negative
stiffness values. In addition, scatter plots of pseudodynamic results reveal that
there is no bias in the orientations of the resultant flexural moments. Therefore it
seems appropriate to make a seismic design based on biaxial bending without bias
in any specific direction. The complicated biaxial hysteretic loops confirm the
difficulty of developing suitable loaddisplacement models for nonlinear dynamic
analysis. These pseudodynamic outputs, such as displacement and hysteretic
responses, can be treated as references for analytical models proposed to simulate
the loaddisplacement relationship. After comparing biaxial with uniaxial hysteretic
loops for both cyclic loading and pseudodynamic tests, it is important to note that
biaxial hysteretic loops show greater stiffness degradation and pinching during
unloading. This reveals that damage caused in one direction weakens the seismic
resistance in the other direction.
Finally, Bechtoula et al. [42] tested eight largescale and eight smallscale cantilever
RC columns under various vertical and horizontal loading patterns. Square and
circular, as well as unidirectional, horizontal displacement patterns were
considered. The axial load was considered both as constant (moderate load and
high load) and variable during the test. From the tests observation, it was clear
that the bidirectional loading had a significant influence on the envelope curves as
well as on the damage progress. The observed damage to largescale columns was
much more severe than that observed for the smallscale columns.
2.2.5 Behaviour of RC columns under biaxial bending
with varying axial load
As widely known, from the axial loadbending interaction diagrams, it is possible
to observe that both yielding and ultimate moments increase with the level of axial
load until the balance point is achieved. Variation in the axial load during an
earthquake can change the strength, stiffness, and ultimate displacement capacity,
as well as all the hysteretic properties of a RC section. Such variations can occur
due to the vertical component of the seismic load, or in the external columns of the
bottom storeys of RC frames, due to the overturning moments which increase the
axial load on one side and decrease it on the opposite side. In fact, the variation in
the axial force during the response cycle may significantly affect the inelastic
response of the columns [32].
26 Chapter 2
Due to testing difficulties, the number of RC column tested under bidirectional
displacement with varying axial load is very reduced and it is always associated
with tests with constant axial load, as summarized in Table 1, thus, the reader few
available results do not allow for solid conclusions to be drawn about coupling
behaviour between biaxial bending and the varying axial force.
The early results of both Li et al. [62] and Low and Moehle [47] evidence similar
effects of axial load variations on uniaxial and biaxial flexure. In particular, it was
found that axial load variation simultaneously with the transverse forces and
deformations leads stiffness and strength increase, while the strength degradation is
larger for higher axial load, and also decreases when the axial load decreases.
The results obtained by Bousias et al. [43] present a strong coupling between the
axial and transverse directions. For the relatively low levels of compressive axial
loads considered therein, deflections induce an axial extension which has a
magnitude roughly proportional to their resultant vector. For lower compressive
loads, the cycling of transverse forces or deflections causes a gradual shortening in
the axial direction under the axial load alone, a ratcheting extension that rapidly
turns into shortening when failure is imminent. As a result of axial/lateral forces
coupling, cycling of the axial force below the balance load causes a ratcheting
increase in the deflection under constant transverse force.
Bechtoula et al. [42] found that the axial load intensity had little effect on the
envelope curve of the secondcycle of the loaddisplacement plot for specimens
under unidirectional horizontal load with constant or variable axial load.
Equivalent viscous damping increased with the increase in axial load, and columns
under variable axial load showed equivalent viscous damping values similar to
those of columns under constant/moderate and constant/high axial load.
2.2.6 Biaxial tests on global bare frame structures
The experimental study of the seismic response of RC frame structures is very
important to understand the behaviour of RC buildings, since this inelastic
response is dependent on each element and on the distribution and participation of
each element in the global response. The available tests in singular elements or
structural subassemblages give an important source of information for
understanding the global structural behaviour; however, the available tests on
concrete frame structures are very limited. Most of the tests performed on frame
structures only focused on planar structures, or threedimensional structures loaded
along one single direction. However, it has been earlier recognized that three
dimensional behaviour may play an important role in the global behaviour of frame
systems, in particular when the torsional response assumes significant importance
due to in plan irregular distribution of the strength and stiffness or mass.
Previous Res
Therefore,
nonlinear
Oliva and
storeys an
table with
shaking t
component
Figu
From the
simultaneo
even thou
than uniax
relatively
The effect
varying ax
to column
influences
increasing
From the
strong col
identifiabl
direction w
in reinforc
deformatio
and damag
in the wea
search and Ba
only 3D m
bidirectiona
d Clough [7
nd four recta
h its longit
able horizo
t of the 195
ure 13 Oliva
frame behav
ous biaxial
gh the mom
xial yield m
to the load
t of the glob
xial load cau
n. Thus, yie
the evolut
the biaxial
analysis of c
lumn axis,
e yielding,
was significa
ced concrete
ons were gre
ge induced
ak axis stiffn
ckground
model testin
al flexure re
0] tested a
angular colu
udinal prin
ontal motio
2 Taft earth
a and Clough
viour observ
moments in
ments along
moments. Th
ding directio
bal torsiona
used by the
elding of ea
tion of the
flexure dem
column beh
parallel to
and the col
antly differe
e members
eater than t
by the stro
ness, with la
ng allow cap
esponse [32]
7/10 scale
umns (see F
ncipal axis
on, the spe
hquake reco
h [70]: Test f
vation, it w
n the colum
g either of t
he nonsymm
on, led to a
al response
e overturnin
ach column
e overall l
mands in the
aviour, it w
o the fram
lumn flexura
ent from the
without an
those along
ong axis mo
arge deforma
pturing the
.
model of a
Figure 13) b
at a 25an
ecimen was
ord, with inc
frame and col
was possible
mns resulted
the column'
metrical stru
a global tor
on the fram
ng moment,
n is reached
ateral stiffn
e corner colu
was observed
me longitudi
al response
e cyclic flex
n obvious yi
the column
tion resulte
ations and l
failures ass
a frame str
by mounting
ngle relative
s subjected
creasing inte
lumn section
to find that
in yielding
's principal
uctural stiffn
rsion in the
me corner c
were differ
d for differe
ness of the
umns.
d that the re
inal axis, e
in the weak
xural respon
ield point,
ns strong ax
d in a consi
ow energy d
sociated wit
ructure with
g it on a sh
e to the ax
d to the N
ensity motio
dimensions
t the existen
g (see Figure
axes were
ness distribu
e frame resp
columns, an
rent from co
ent steps, w
e frame sy
esponse alon
exhibited re
k (or transv
nse normally
even thoug
axis. The yie
iderable dec
dissipation.
27
th the
h two
haking
xis of
N69W
ons.
nce of
e 14),
lower
ution,
ponse.
nd the
olumn
which
ystem,
ng the
eadily
verse)
y seen
gh the
elding
crease
28
Fig
The
const
earth
confi
from
centr
has a
the o
of ps
retro
levels
direc
reach
absor
direc
of th
The
the s
torsio
eccen
Final
litera
dime
biaxi
what
gure 14 Oli
SPEAR st
tructions in
hquake resi
guration is
3 to 6m; t
re of mass.
a crosssecti
others along
seudodynam
ofitted confi
s: 0.02, 0.1
ction, was g
hed in this
rbed energy
ction on the
he absorbed
interstorey
specimen, d
onal irregu
ntricity and
lly, it is wo
ature were
ensional load
ial behaviou
t regards the
iva and Cloug
tructure wh
n Southern
istance [71]
nonsymme
there is a b
Eight colum
ion of 250x7
g the Y dire
mic (PsD) te
igurations.
15 and 0.20
lobally stro
direction w
y; on the con
e second floo
energy at e
drifts at ea
due to torsio
larity on t
relatively lo
orth referrin
developed
dings, so th
ur of colum
e analysis of
gh [70]: Mom
c
hich is a t
European
]. The str
etric in two
balcony, shi
mns have a
750mm, wh
ction. The t
ests on the
Each set c
0g [72]. The
nger. Thus,
were larger,
ntrary, much
or. These fi
each floor. T
ach storey w
onal effects.
the column
ow levels of
ng that mos
aiming to
hey were n
ns. So, the
f the column
ment vs base r
column
threestorey
countries,
ructure is
o directions
ifting the c
a square 250
hich makes i
test program
specimen in
consisted o
e Y directio
as expecte
in agreeme
h larger disp
indings were
The most af
were differen
. The tests
n drifts, ev
f excitation
st of the te
o study th
not particula
conclusions
n response a
rotation resu
y building
without s
regular in
s, with two
centre of st
0x250mm cr
it much stif
m consisted
n the origina
of three tes
on, parallel
d, the level
ent with th
placements
e confirmed
ffected level
nt for each o
highlighted
ven for a l
[71].
esting camp
he building
arly focused
s from thos
at local leve
lts from a rec
representat
pecific prov
elevation,
obay frame
iffness away
rosssection,
ffer and stro
of three dif
al and in tw
sts at diffe
to the stro
s of intersto
he global va
were reache
d by the tim
l was the se
of the nine
d the strong
limited leve
aigns availa
s response
d in the stu
se research
el is limited.
Chapter 2
ctangular
ive of old
visions for
the plan
es spanning
y from the
, while one
ronger than
fferent sets
wo different
erent PGA
ong column
orey shears
alue of the
ed in the X
me histories
econd level.
columns of
g effects of
el of plan
able in the
to three
udy of the
projects in
Previous Research and Background 29
2.3 Numerical modelling of biaxial bending of RC
columns
2.3.1 Introduction
Present seismic design recommendations intend that buildings should respond
elastically only to small magnitude earthquakes and, therefore, the need of
adequate nonlinear models to account for the hysteretic behaviour of RC
structures is of paramount importance. In terms of earthquake response, buildings
may present either a simple or a very complex behaviour. The differences are
dependent on layout and other architectural features, type of structural system and
presence of irregularities. Consequently, the structural model used in the analysis
should be dependent on the building type, the design process and the level of
seismicity [73]. Normally, nonlinear analysis can reproduce the influence of the
structural system geometry and of the mass and stiffness distribution, particularly
in structures with irregular stiffness and mass variation in plan and/or in elevation.
The model types can be divided into different categories, namely global models,
microscopic models and discrete finite element (member) models. In global models,
the nonlinear response of the structure is concentrated in selected degrees of
freedom having hysteretic characteristics that can be used to simulate a group of
structural elements, or even the entire structure. This type of model can be used
for a preliminary evaluation of the interstorey drifts and ductility demands [18].
For microscopic models, the RC element is discretized into a larger number of
elements, with different constitutive models for the materials (concrete and
reinforcing steel), and for their possible interaction through bonding, thus allowing
accurate representations of the element geometry and a full knowledge of all the
elements points of stress and strains [32]. However, and even considering the
computational advances in recent decades, the use of this type of model is
prohibitive for multistorey structures. Therefore, it can be used for the analysis of
individual RC elements or subassemblages to study local phenomena, or to
support the calibration of more simplified models [38].
In between these two types of modelling strategies are the membertype models,
that appear to be the best available compromise between accuracy of structural
response, computational cost and simplicity of use [38]. In these models, each model
element has a correspondence to an element of the structure. It is not possible to
model all the individual aspects of the structure as in microscopic models, but they
do allow a close representation of the key features of the structural behaviour [74].
This is the modelling strategy is appropriated for the seismic design and assessment
procedures as part of common engineering practice, where some kind of nonlinear
30
analy
levels
tradi
The
frame
requi
differ
behav
2.3.
A va
behav
formu
[76].
The
lump
at th
eleme
decad
with
bend
The
there
ysis is requi
s of accurac
itionally use
member ty
es, because
irements an
rent memb
viour.
.2 Nume
ast number o
viour of
ulations to
Figure 15
lumpedplas
ped, at pred
he centre of
ents [77]. S
des, starting
the conside
ding interact
lumpedpla
efore, so ha
ired, either
cy. At the
ed: lumpedp
pe models a
e they pres
nd the optim
ber type m
erical mo
of models h
beamcolum
distributed
Element mod
sticity mode
determined s
the plastic h
Several lum
g with the f
eration of t
tion [60, 81]
asticity mod
ave inheren
static or dy
level of ind
plasticity an
are the mos
ent the be
mum possib
models whic
odelling
have been de
mn member
plasticity fo
del approach
beam
els assume t
sections. No
hinge zone,
mpedplastic
first studies
the bending
and bendin
dels are a
nt deficienc
ynamic, usi
dividual ele
nd distribute
st widely us
est comprom
ble outcome
ch are abl
strategie
eveloped to
rs, ranging
formulations
hes for nonlin
m\columns
that nonlin
onlinear be
which is ge
city models
by Clough
g and axial
ng/shear int
simplificat
cies. The a
ing frame e
ments, two
edinelastici
sed for non
mise betwee
e. The next
e to repre
es at the
represent t
g from co
s based on f
near numeric
near behavio
haviour is a
enerally loca
have been
et al. [78],
force intera
eraction [80
ion of the
assumption
lements wit
main appr
ity models [7
linear anal
en the com
t section pr
esent biaxia
e elemen
he material
oncentrated
fnite elemen
al modelling
our is conce
assumed to
ated at each
n proposed
and develop
action [79, 8
0, 82].
real behav
that the
Chapter 2
th different
roaches are
75].
lysis of RC
mputational
resents the
al bending
nt level
l nonlinear
plasticity
nt methods
of RC
entrated, or
be located
end of RC
in recent
ped further
80], biaxial
aviour and,
damage is
Previous Research and Background 31
concentrated in the plastic hinge, does not allow using this type of models for some
RC elements where the behaviour is not governed mainly by bending, like shear
walls. In this cases the shear behaviour can have a significant contribution, major
inclined shear cracks are expected to occur in the midheight element, and the
consideration of plasticity concentrated in the element ends is not an accurate
assumption [83]. The lumped models simplify some aspects of hysteretic behaviour
of RC members, because the characteristic section behaviours are defined a priori.
The main proposals can be found in the literature as described by Otani [40],
Fardis [74], Filippou and Fenves [84].
Another modelling strategy approach at the element level is associated with the
distribution of nonlinearity along the element, by modelling it with a certain
number of controlling sections, including the nonlinear behaviour which is
integrated to obtain the global nonlinear response of the element. This concept
was first introduced by Otani [85] and the major advantage of these models is the
nonexistence of a predetermined length where the inelasticity can occur, because
all the sections can have incursions in the whole response range (linear and non
linear). While this approach is a closer approximation to reality, it also requires
more computational capacity [86]. An exhaustive review describing these
formulations can be found in the work of Taucer et al. [18], Spacone et al. [87] and
Arde [38].
At the section level, different strategies can be used. The most commonlyused in
the lumped plasticity models is to describe the sectional nonlinear behaviour
through a momentcurvature skeleton curve, associated with hysteretic rule to
describe the global section nonlinear behaviour. For the distributed plasticity
models, the common method for computing the section response represent through
fibre models, where the crosssection is subdivided into concrete fibres and steel
fibres, the material properties being defined by uniaxial inelastic stressstrain law
for each material type.
2.3.3 Modelling biaxial flexure with axial force
The present knowledge of RC member modelling under cyclic biaxial bending with
axial load is still far behind that of uniaxial bending; however, the response of RC
buildings to both earthquake components is very important, especially the
behaviour of vertical elements that are subject to biaxial bending. In fact,
international guidelines already accept nonlinear analysis (static and dynamic) for
the evaluation of the seismic behaviour of existing structures and design of new
ones, such as the ATC 40 [88], BSSC [89, 90], and Eurocode 8 [17]. However, the
available biaxial models are not mature enough to be incorporated into code
standards, as is now with the case for uniaxial simplified global models.
32
Diffe
cyclic
mode
Load
mode
classi
hyste
dama
The
and a
the g
relati
the f
[18]
defor
types
large
curva
Fig
Lai e
biaxi
conce
the d
discr
sprin
a sop
eithe
tool
struc
with
rent model
c behaviour
els is presen
ding State
els [18, 87, 9
ical plastici
eresis model
age models
fibre model
axial effect.
geometric ch
ionship of th
fibres respon
(see Figure
rmations [75
s of modell
er than tho
ature relatio
gure 16 Fib
et al. [105,
ial bending
entrates the
description o
etizes the
ngs. These ty
phisticated
er a skeleton
for threed
ctures [106].
the consid
ling strateg
r of RC elem
nted in the
e of the Ar
91], other a
ity [92], Mr
lling [20, 96
[51, 100, 10
s are able t
In these m
haracteristic
he section is
nse, which f
e 16). Adv
5, 91, 102] a
ling require
ose based o
ons to descr
bre element:
106] have p
behaviour a
e nonlinear
of stressstr
crosssectio
ypes of mod
approach a
n curve or t
dimensional
Over the y
deration of
gies have b
ments with a
e CEB Rep
rt Report
nalytical m
roz multisur
698], bound
1].
o capture c
models the el
cs of which
s not specifi
follow the u
vanced fibre
and the bon
computer
on simplifie
ibe the sect
distribution o
fibres (ada
proposed a
and the axia
behaviour i
rain propert
n into fou
del are very
and simplifi
threedimen
inelastic a
years, the m
more sprin
been propos
axial force;
port N220,
[32] and b
models are av
rface plastic
ding surface
coupling effe
lement is su
are their lo
ied explicitly
uniaxial stre
e models a
nd slip [103]
time, stora
ed macroel
ion behavio
of control sec
apted from [1
triaxial spr
al force vari
into the elem
ties of steel
ur effective
accurate, b
fied models.
nsional yield
analysis an
model has be
ngs, and di
ed for the
a detailed r
RC Fram
by Fardis [
vailable, foll
city [60, 93
e plasticity
ects in the t
ubdivided in
cation and
y, but is der
essstrain rel
are now ab
. Despite th
age and inp
ements wit
our [19].
ctions and se
104])
ring model
iation (see F
ment ends,
and concret
steel sprin
but try to b
. There is
d surface, m
nd design o
een subject t
fferent cons
simulation
review of th
mes under E
74]. Besides
lowing the c
, 94], Bouc
[19, 81, 99]
two bending
nto longitud
area. The c
rived by int
lation of eac
ble to cons
heir great po
put prepara
th empirica
ction subdivi
that can sim
Figure 17). T
which is con
te. The orig
ngs and fiv
ridge the ga
no need to
making this
of reinforced
to some imp
stitutive law
Chapter 2
of biaxial
he available
Earthquake
s the fibre
concepts of
cWen [95],
or lumped
g directions
dinal fibres,
constitutive
tegration of
ch material
sider shear
ower, these
ation much
al moment
ision into
mulate the
This model
ntrolled by
ginal model
ve concrete
ap between
o construct
a practical
d concrete
provements
ws for the
Previous Res
springs [10
results wh
buildings [
As for the
distance o
current lo
obtained t
were devel
presented
response o
plasticity
additional
of the axia
responses
accurate t
important
The hyste
not been
with biax
behaviour
Kunnath a
account f
viscoplasti
uniaxial b
differentia
combinatio
controlled
strength d
biaxial be
search and Ba
07109]. Th
hen applied t
[109] subject
e so called b
of the curre
oading direc
through a s
loped by Sfa
accurate res
of a reinfo
biaxial flex
features su
al load varia
of slender c
than fibre
parameters
eresis model
greatly dev
xial bending
(Takedaty
and Reinhor
for the effe
ic principle
behaviour d
l equation,
on of an el
by two pa
deterioration
ending inter
ckground
e applicatio
to isolated R
ted to biaxi
Figure 17
bounding su
ent generali
ction. At t
section analy
akianakis an
sults for cyc
orced concre
xural mode
uch as stiffn
ation, in ord
olumns. Bou
models, h
s of the seism
s for repres
veloped. Th
g interactio
ype) has b
rn [96] prop
ect of biax
s. The und
described by
with the m
lastic force
arameters w
n. Park et
ractions, us
on of the L
RC element
ial loadings.
7 Triaxial s
urface mode
ized stress
the section
ysis procedu
nd Farfis [8
clic uniaxial
ete buildin
el [60, 93],
ness degrada
der to achie
usias et al.
however it
mic respons
senting the
he developm
on, similar
been consid
posed a two
xial bendin
derlying sch
y Bouc [11
model expre
and a plas
which separa
al. [111] p
sing an equ
ais spring
ts [39, 62] a
spring model
els, the inel
point to th
n level, the
ure. The fir
1, 99] and B
l or biaxial t
g [99]. Ba
, Gogal an
ation, shear
ve an accur
[19] state th
provides d
e.
biaxial beh
ment of a p
to typical
dered to be
dimensiona
ng interacti
heme of th
10], which
essing the r
stic force. T
ately contro
roposed a h
uivalent line
model prov
nd also whe
[18]
asticity is c
he bounding
e bounding
rst bounding
Bousias et a
tests, as wel
ased on th
nd Ghobara
r deformatio
rate model f
hat this typ
direct contr
aviour of R
henomenolo
l models u
e close to
l forcedefor
ion in colu
his model i
involves th
restoring for
The hystere
ol stiffness
hysteretic m
earization p
vides an acc
en applied t
controlled b
ng surface i
surface ca
ng surface m
al. [19, 43], w
ll as the dyn
he basic lu
ah [94] inc
on and the
for predictin
e of model i
rol of the
RC columns
ogical 3D m
used in un
impossible
rmation mo
umns, base
is based on
he solution
rce as the
etic behavio
degradation
model to in
process to
33
curate
to RC
by the
n the
an be
models
which
namic
umped
cluded
effect
ng the
is less
most
have
model,
niaxial
[19].
del to
ed on
n the
of a
linear
our is
n and
nclude
study
34 Chapter 2
random vibration of hysteretic systems. Romo et al. [20] proposed an extension of
existing piecewise linear (PWL) uniaxial hysteretic models, by adapting the biaxial
structural interaction form of the BoucWen biaxial hysteretic model to couple the
two loading directions. The model was demonstrated through the analytical
simulation of available biaxial experiments on RC columns, with accurate results;
however, additional research was still found necessary to define objectively the
interaction scaling factor which rules the coupling of the two loading directions and
also, to capture more accurately the strength capacity decrease relative to the
virgin loading curve. The main advantages in using these models are the good
results obtained, associated with the lower computational cost and the
implementation and application simplicity.
Marante and FlrezLpez [51, 100] proposed a lumped damage model based on
modelling of the biaxial bending of RC frames within a new framework called
lumped damage mechanics, which assumes that the damage processes are
uncoupled in both orthogonal directions. The damageuncoupling assumption
allows the model parameters and the crack strength terms to be computed in a
very simple way by considering four monotonic loadings. The resulting model is
conceptually as simple as the twodimensional one and provides good enough
results for engineering purposes.
Mazza and Mazza [101] developed a lumped plasticity model for the nonlinear
analysis of threedimensional reinforced concrete frames subjected to bidirectional
ground motion. The frame members are idealized by means of two parallel
elements, one elasticperfectly plastic and the other linearly elastic, assuming a
bilinear moment curvature law. An interaction surface is considered for the end
sections of each frame member, relating the axial force and biaxial bending
moments. The model was only compared with refined fibre model results, providing
satisfactory simulation of the flexural hysteretic behaviour of RC elements with
axial forcebiaxial bending moment interaction. The lumped plasticity models are
relatively simple and therefore, can be efficiently incorporated into the nonlinear
dynamic analysis of complex multistorey RC structures.
2.4 Final remarks
The response of RC columns is recognized as a very important topic that should be
taken into consideration for building structures in earthquakeprone regions. The
response of RC building columns to earthquake actions deals in general with three
dimensional responses, due to the random characteristics of the earthquake
direction and to the building irregularities.
Experimental research work on the inelastic response of RC members under
compression axial force and biaxial lateral cyclic bending loading conditions is
Previous Research and Background 35
currently very limited. Uncertainties concerning the relationship and combination
of the two orthogonal horizontal loading paths, associated with the complexity of
the experimental setup, certainly justify this lacuna.
From the analysis of the available experimental work, it was possible to verify that,
besides the expected significant influence of axial loads on the hysteretic response
of columns, the bidirectional transversal load cycles are responsible for an
increasing in the stiffness and strength degradation, when compared to the uniaxial
response. In addition, the failure mechanism of RC columns is found to be very
dependent on the loading path, which strongly affects both the ductility and energy
dissipation capacity of the columns. On the other hand, there is some experimental
evidence that plastic hinge lengths tend to be stable at around theoretical values
and are not strongly affected by bidirectional loading.
From the available modelling methodologies the membertype models are found to
be able to ensure accurate representations of the key features of RC elements
behaviour. The existence of complex and irregular RC structures implies that
threedimensional analysis, including nonlinear behaviour of RC elements should
be adopted so as to simulate the response of columns subjected to biaxial bending
and axial loading. Fibre models provide an accurate methodology for capturing the
complex biaxial behaviour under constant and varying axial loads. However, the
use of more simplified models, such as bounding surface models, spring models or
lumped plasticity models, presents a major advantage in for representation of
global seismic response with less input preparation and less computer time and
storage.
Chapter 3
Test Program
3.1 Introduction and objectives
The experimental studies are considered crucial to understand the behaviour of
structural elements or global structures and help the development of numerical
models. As previously referred in Chapter 2 the number of test results on RC
columns subjected to biaxial bending and axial load is not so extensive when
compared to those on uniaxial bending, as so it can be considered insufficient to
truly understand the interaction between the orthogonal directions.
The experimental work presented in this Chapter is part of a large testing
campaign undertaken by the Laboratory of Earthquake and Structural
Engineering (LESE), of the Faculty of Engineering of Porto University (FEUP),
for the study of RC columns (of buildings and bridges) under horizontal cyclic
loadings [112118]. The main objective of this experimental campaign was to study
the cyclic behaviour of RC columns, under uniaxial and biaxial horizontal cyclic
loadings in order to: i) improve the lack of experimental results in RC columns
under biaxial bending; ii) contribute for understanding the effects of biaxial
bending when compared to the uniaxial bending; iii) test the influence of different
biaxial load paths; iv) calibration of existent numerical models and development of
simplified numerical models for the simulation of nonlinear behaviour of RC
elements.
The experimental program included the construction and testing of four types of
rectangular RC columns, with different geometric cross section and longitudinal
reinforcement detailing, tested under uniaxial and biaxial lateral loading with
constant axial force, amounting to a total of 24 specimens. The present chapter
first describes the specimens characteristics and construction; subsequently it
38
prese
and
adop
3.2
3.2.
This
speci
conte
recta
unde
teste
contr
For
was a


ents the ma
the conside
ted instrum
Specim
.1 Introd
Section des
imen detaili
ext, and fo
angular RC
er uniaxial l
d under bia
rolled condit
each colum
adopted, wh
$$ takes
assumed
uniaxial
the value
## repr
Figure 18
aterials me
ered horizon
mentation sch
mens an
duction
scribes the f
ing; the co
or the pur
columns we
loading (in
axial loading
tions.
mn specimen
here:
s the value
herein as
tests on th
e 12 for b
resents the r
8  General sc
chanical pr
ntal load pa
heme are pr
nd cons
four types of
nstruction
rpose of th
ere built. Fo
the strong
g, all of them
n test, the
01 for u
the X dire
he weak dire
iaxial tests;
reference nu
cheme of the
roperties, th
aths. Finall
resented.
structio
f the column
process is
he study h
or each typ
and weak d
m with cons
following g
uniaxial tes
ection, see
ection (WE
umber of the
column spec
he constant
ly, both the
on
ns construct
also briefly
herein addr
e series, two
directions),
stant axial f
general desig
st on the s
Figure 18)
E or Y dire
e column spe
cimens and te
axial loadi
e test setu
ted for testi
documente
ressed, four
o columns w
while the o
force and dis
gnation PB
strong direc
), the value
ection, Figu
ecimen.
esting directio
Chapter 3
ing aspects
up and the
ing and the
ed. In this
r types of
were tested
others were
splacement
B$$N##
ction (NS,
ue 02 for
ure 18) and
ons
Test Program 39
3.2.2 Prototype columns: geometry and section detailing
The specimens were constructed at full scale, assuming that the inflection point of
the column deflected shape is located at the column midheight. All the column
specimens are 1.70m high and were cast on a strong square RC foundation block
with 1.30x1.30m
2
in plan dimensions and 0.50m high. Four holes are arranged at
the foundation block in order to fix the specimen to the laboratory strong floor
with prestressed steel rods. This process avoids sliding and overturning of the
column footing during the test.
Figure 19, shows the general layout of the geometry and section detailing of the
RC specimens. The columns N01 to N04, are characterized by a rectangular
0.20m x 0.40m cross section with six 12mm diameter bars for the longitudinal
reinforcement; columns N05 to N08, N17, N21 and N22 have 0.30m x 0.40m
rectangular cross section with longitudinal reinforcement composed by ten 12mm
diameter bars columns N09 to N12, N18, N19 and N20 have 0.30m x 0.50m
rectangular cross section with fourteen 12mm diameter bars and
columns N13 to 16, N23 and N24 consist of 0.30m square cross section with eight
longitudinal 12mm diameter rebars. For all the columns the transversal
reinforcement is made of 6mm diameter stirrups spaced at 150mm and the concrete
cover was taken 3cm.
Although typical RC building columns have lap splices near the base, which
potentially influences the columns response to cyclic loadings (such as earthquake
induced ones), for this experimental testing program columns were built with
continuous longitudinal reinforcement bars, i.e. without lap splices. This deliberate
option is justified because the main objective of this experimental research is to
study the flexural behaviour of biaxially loaded RC columns and the consideration
of the lap splices would induce additional variables to the problem, thus making
more difficult the understanding of the global column behaviour. This strategy has
been also followed in many previous works by other authors.
40 Chapter 3
a)
b)
Figure 19  RC column specimens dimensions and reinforcement detailing:
a) crosssections details; b) specimen dimensions and general scheme of the reinforcement
layout
3.2.3 Construction
The specimen construction was rigorously supervised in order to achieve the design
specifications in terms of materials, structural geometry and reinforcement
detailing. All the RC specimens were built on a construction site and traditional
workmanship was adopted in order to convey the common construction practice as
close as possible. The concrete was provided from a readymixed concrete plant and
vibrated in the casting phase with a usual vibrating device.
Figure 20 shows pictures with relevant details of the different phases of the
construction of columns. After the construction the geometry of the specimens was
verified. In the first series (columns N01N04) same problems were detected with
the columns verticality. Columns were cast in a single stage, first the footing and
then the column, without cold joints. Each series of the four column types was cast
in different days. Although for columns N05N24 the concrete specification and the
readymixed concrete plant were the same, slight differences can be expected in the
obtained concrete properties.
N01N04 N05N08
N09N12 N13N16









 



 
Test Program
Figure 2
connection
3.3 Ma
The mate
concrete, c
N24 (NPE
Materials
steel reinfo
in the colu
m
20 Construc
n; c) stirrups
nee
aterial
erials consid
class C35/4
EN 2061 [1
tests were
orcement) in
umns; the ob
ction of the c
s in place; d)
edle; f) specim
propert
dered at th
5 for colum
119]), and r
made on sa
n order to c
btained resu
c)
columns: a) f
column cast
mens after re
ties
he specime
mns N01 to
reinforcing
amples taken
characterize
ults are pres
a)
e)
formwork and
ing; e) concre
emoving the f
en design p
N04 and a
steel of clas
n during the
the real pro
sented and a
d column foot
ete vibration
framework
phase consi
C25/30 for
ss S400 for
e constructi
operties of t
analysed nex
ting; b) colum
n with a vibra
isted on re
columns N
all the colu
ion (concret
the material
xt.
b)
41
mn
ating
egular
N05 to
umns.
e and
l used
d)
f)
42 Chapter 3
3.3.1 Concrete
The C35/45 concrete considered in the design of columns N01 to N04, was
characterized by maximum aggregate dimension of D22 (in mm) and slump class
S3, while for the C25/30 concrete (assumed for columns N05 to N24) the maximum
aggregate dimension of D31 (in mm) and slump class S2 were adopted in
accordance with the NPEN02061 Portuguese Standard [119].
The nominal properties of the used concretes are specified in Table 2 according to
the Eurocode 2 [120]. During the casting phase of the first series (column N01N04)
150mm cubic concrete samples were taken, while for the other columns series
(N05N24) six cylindrical concrete samples (30cm high and 15cm diameter) were
taken from each casting phase. Table 3 summarizes the results of the corresponding
compression tests of tree column series samples and of the diametric compression
tests of one sample for each column series (to estimate the tensile strength),
corresponding to different casting phases. These material tests were performed by
the actual time of the columns experimental tests.
Table 2 Nominal values to the concrete considered in the design
Specimen
group
Class of
concrete
Compressive ultimate strength [MPa]
f
ck,cylinder
f
ck,cubic
N01N04
C35/45
35
45
N05N24
C30/35
30
35
Table 3 Results from compression tests on concrete specimens
Specimen group
Compressive ultimate strength
[MPa]
Tensile strength
[MPa]
Sample 1 Sample 2 Sample 3 Average
Columns N01N04
*
44.89 51.08 49.08 48.35 N/A
Columns N05N08
**
21.07 22.12 21.02 21.40 2.35
Columns N09N12
**
24.73 23.82 24.66 24.39 2.57
Columns N13N16
**
20.89 22.65 21.57 21.70 2.98
Columns N17N18
**
Columns N23N24
**
36.26 36.63 36.11 36.30 3.24
Columns N19N22
**
43.63 43.04 42.78 43.14 3.28
*
cubic concrete samples
**
cylindrical concrete samples
Concerning the test results of the concrete specimens, the 1
st
series (column N01
N04) have confirmed the expected results, and have a mean value of the concrete
cubic strength of 48.35MPa, thus indicating a good concrete in agreement with the
expecting value. The results for the concrete cylindrical specimens of the 2
nd
to the
4
th
series (columns N05N016) have shown a poor quality concrete with a mean
Test Program 43
value of the compressive strength between 21 and 24MPa. These values are clearly
below than the expected ones for to the ordered concrete, but this fact was duly
taken into account in the result analysis so as to not affect the main objectives of
the present study. For the columns N17 to N24 the results are in a better
agreement with the expected values, although for the 5
th
and 6
th
series the
compressive mean strength is about 20% below what should comply with the
ordered C35/45 class.
Regarding the concrete tensile strength, considering that only one test was
performed for each casting phase, the results can be considered acceptable and are
in agreement with the expected results representing around 10 to 15% of the
compressive strength.
3.3.2 Reinforcement steel
The steel class S400 was considered for both the longitudinal and transversal
reinforcement steel of all columns; the nominal values of mechanical properties are
presented in Table 4, according to the REBAP [121] Portuguese standard.
Table 4 Nominal mechanical properties of steel S400 [121]
Young modulus
E [GPa]
Yielding strength
f
sy
[MPa]
Ultimate strength
f
su
[MPa]
Ultimate strain
su
[%]
200 400 460 14
For the columns construction, three different lots of reinforcement steel were used,
namely one for columns N01N04, another for columns N05N16 and another for
columns N17N24; therefore, three samples of 12mm diameter bars (longitudinal
steel) were taken from each lot and tested according to the NPEN100021 [122].
The stressstrain plots are presented in Figure 21, confirming the expected type of
steel behaviour.
The relevant obtained results are summarized in Table 5, in terms of Young
modulus, yielding strength, ultimate strength and ultimate strain, for each sample
and for the corresponding average value of each steel lot.
The mechanical properties of the steel used in the specimens, are in a good
agreement with the nominal values (Table 4), although the mean values are
slightly higher than the expected ones.
44 Chapter 3
Figure 21 Steel stressstrain curves: a) columns N01N04; b) columns N05N16
c) columns N17N24
Table 5 Results from the tensile tests on steel samples
Specimen
group
Sample
Young modulus
Yielding
strength
Ultimate
strength
Ultimate
strain
E [GPa] f
sy
[MPa] f
su
[MPa]
su
[%]
Columns
N01N04
1 197.75 435.77 543.80
22.4
2 195.00 435.20 547.50 20.15
3 191.23 426.93 542.68 21.74
Average 194.66 432.63 544.66 21.43
Columns
N05N16
1 204.86 428.27 548.87 22.97
2 207.67 431.88 554.39 23.36
3 197.85 428.91 549.88 20.88
Average 203.46 429.69 551.08 22.40
Columns
N17N24
1 177.76 451.23 579.95 19.4
2 201.31 449.29 573.89 21.39
Average 189.53 450.26 576.92 20.39
3.4 Axial load
Concerning the axial load, it was considered constant during the tests. The axial load
values and the normalized axial stress (calculated with the axial force and the average
material properties described in Section 3.3.1) are summarized in
Table 6.
The normalized axial stress (also known as the axial load ratio) considered in the
tests corresponds to typical values for current reinforced concrete buildings with 3
to 4 storeys. The Eurocode 2 [120] points out the minimum value of 0.1 for the
normalized axial stress in terms of characteristics values of both the loading and
0 5 10 15 20 25
0
100
200
300
400
500
600
S
t
r
e
s
s
(
M
P
a
)
Strain (%)
sample 1
sample 2
sample 3
0.23% 3.57%
0 5 10 15 20 25
0
100
200
300
400
500
600
3.04% 0.22%
S
t
r
e
s
s
(
M
P
a
)
Strai n (%)
sampl e 1
sampl e 2
sampl e 3
0 5 10 15 20 25
0
100
200
300
400
500
600
2.32% 0.22%
S
t
r
e
s
s
(
M
P
a
)
Strain (%)
sample 1
sample 2
a) b)
c)
Test Program 45
the material properties, which corresponds to smaller values when considering the
mean values rather than characteristic ones.
3.5 Horizontal displacement paths
One of the main objectives of the present experimental study was to compare the behaviour
of RC columns subjected to uniaxial displacements paths and the effect of the different
biaxial displacement paths in rectangular columns. For columns N17 to N24 different levels
of axial force were considered taking into account the objectives of the study, particularly
the effect of the axial force level in the column response. Thus, for each column type
different levels of normalized axial stress were adopted. For the main study and for each
column type, uniaxial and biaxial tests were performed as shown in
Table 6 (series 1 to 4).
In order to characterize the response of the column specimens, cyclic lateral
displacements were imposed at the top of the column with increasing demand
levels. Three cycles were applied for each of the following peak
displacements: 3, 5, 10, 4, 12, 15, 7, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75,
80mm. However, several times it was not possible to achieve the cycles of larger
amplitude due the damage reached in the specimens. This procedure allows for easy
understanding of the columns behaviour, and consequently, for the comparison
between tests and the development and calibration of numerical models. The
cycles repetition for each displacement demand allows obtaining information to
better understand the stiffness and strength degradation of the column, which is
relevant also for the calibration of numerical models.
For the biaxial tests four different patterns were considered (see Figure 22), namely one
where the displacement cycles are applied alternately in the two horizontal directions
(cruciform path type 3), one with a rhombus shape (path type 4), one expanding square
centred in the origin (path type 5) and finally a circular load path (type 6). The same load
paths were used for studying the axial force level influence in the Series 5 and 7 (
Table 6). In order to study the influence of cycles repetition for each displacement
demand and of the smooth increase of the load path, the displacement patterns
geometry were maintained in Series 6 but with less peak displacements and without
cycles repetition for each displacement level; thus, in that case, only the following
peak displacements (mm) were considered: 3, 5, 10, 4, 12, 15, 7, 20, 30, 40, 50, 60,
70mm.
46 Chapter 3
Type 1 Type 2 Type 3
Type 4 Type 5 Type 6
Figure 22 Type displacement paths
Table 6 Test series identification, specimens main data and displacement paths
Series Column
Geometry
[cmxcm]
f
cm
[MPa]
N
[kN]
N/(A
c
.
f
cm
)
Displacement
path type
1
PB01N01
20x40 48.35 170 0.04
1
PB02N02 2
PB12N03 3
PB12N04 4
2
PB01N05
30x40
21.40 300
0.12
1
PB02N06 2
PB12N07 4
PB12N08 5
PB12N17 36.30 510 6
3
PB01N09
30x50
24.39 300
0.08
1
PB02N10 2
PB12N11 4
PB12N12 5
PB12N18 36.30 440 6
4
PB01N13
30x30 21.57 210 0.1
1
PB01N14 4
PB01N15 5
PB01N16 6
5
PB12N19
30x50 43.14
600 0.09 4
PB12N20 300 0.045 4
6
PB12N21
30x40 43.14 620 0.12
4
PB12N22 5
7
PB12N23
30x30 36.30 650 0.2
4
PB12N24 5
X
Y
X
Y
X
Y
X
Y
X
Y
X
Y
Test Program
3.6 Te
3.6.1 Te
Figure 23
campaign
the lateral
with +/1
which two
were used
laboratory
overturnin
m
est setu
est setu
and Figur
on RC colu
l loads (one
00mm strok
o steel react
d. Both th
y strong floo
ng (of the sp
Figure 23
Figure
up and
up
e 24 shows
umns. The s
with 500kN
ke) and a ve
ion frames
e specimen
or with pre
pecimen or t
 Testing set
24 Testing
Actuator
500kN with +/150m
Steel
Specim
S
t
e
e
l
r
e
a
c
t
io
n
f
r
a
m
e
Vertical Actuator
700kN
instrum
s the setup
system incl
N with +/1
ertical 700k
(lateral and
ns and the
estressed st
the reaction
tup at LESE
setup at LES
mm
reaction frame
men
e
mentati
adopted fo
udes two ho
50mm strok
kN actuator
d vertical) a
reactions
eel bars in
frames) du
E laboratory:
SE laboratory
Actuator
200kN with +/100mm
ions
or the expe
orizontal ac
ke and the o
to apply th
and one late
frames we
order to a
ring the tes
schematic lay
y: general vie
R
e
a
c
t
i
o
n
w
a
l
l
erimental te
ctuators to
other with 2
he axial loa
eral reaction
ere fixed to
avoid sliding
sts.
yout
ew
47
esting
apply
200kN
ad, for
n wall
o the
g and
48
As st
loadi
actua
durin
the t
force
the t
to di
the fo
The
press
force
really
syste
The
throu
routi
Data
cells,
ampl
tated before
ing was cyc
ator remain
ng the test,
top column
effects. Ho
wo horizont
isplace later
orces read b
Figure 2
hydraulic s
sure, thus e
is monitore
y imposed lo
em occurs.
horizontal
ugh PXI co
ines based o
a acquisition
LVDTs (
lified analog
e, constant
cled under d
s in the sam
a special de
section an
owever, in o
tal direction
rally) and t
by the load
5 Axial loa
system of th
ensuring a c
ed during th
oads on the
actuators
ontroller sys
on the LabV
n and signa
(Linear Var
gical or digit
axial load w
displacemen
me position
evice consist
nd the actua
rder to mea
ns are conne
the correspo
cells of the
ad actuator an
he vertical
constant ax
he test in or
specimen, i
control and
stems run b
VIEW softw
al conditioni
riable Disp
tal sensors [
was applied
nt controlled
while the c
ting of two
ator, in ord
asure these
ected to the
onding mea
horizontal a
nd sliding ste
actuator w
xial force du
rder to hav
in case any
d the data
by home d
ware platfor
ing cards p
lacement T
112].
d to the colu
d conditions
column spec
sliding steel
der to mini
small frictio
upper plate
asured force
actuators.
eel plates at
was designed
uring each t
e complete
unexpected
acquisition
developed co
m (www.ni
provide dire
Transducers)
umns while
s. Since the
cimen latera
l plates exis
mize spurio
on forces, lo
e (that is ex
s are subtra
the top colum
d to keep co
test. Noneth
information
malfunctio
n are both
ontrol and
.com) (see F
ct readings
) and othe
Chapter 3
the lateral
axial load
ally deflects
sts between
ous friction
oad cells in
xpected not
acted from
mn
onstant oil
heless, this
n about the
oning of the
performed
acquisition
Figure 26).
from load
er types of
Test Program
3.6.2 In
Since the g
predomina
tests had
first two u
designed f
(LVDTs)
the instrum
Figure 27
the latera
displaceme
curvature.
considering
perpendicu
In order
orthogonal
LVDTs (s
from the
from the c
Due to th
deformatio
error in th
during the
readings w
LVDTs loc
m
Figur
nstrumen
global defor
ant flexural
to take thi
uniaxial tes
for the biax
that would
mentation la
shows the
al displace
ents in seve
For the u
g only one
ular to the t
to measure
l directions
see Figure
column bas
column spe
he column d
ons, the inc
he measured
e analysis of
with the v
cated at the
e 26 PXI c
ntation
rmation of t
deformatio
is fact into
sts with a v
xial tests, d
d be require
ayout evolv
final adopte
ments at
eral points s
uniaxial tes
LVDT for m
test one, in
e the colum
(X and Y)
28) fixed a
se section. T
cimen and
deformed sh
clination int
d displacem
f the experim
vertical disp
e column fac
ontroller and
hese slender
ons, the req
account. T
very detaile
due the larg
ed. Therefor
ved to the fin
ed instrume
several he
strategically
sts the inst
measuring t
order to eva
mn specimen
, there was
at different
The LVDT
attached to
hape achieve
troduced in
ments. Howe
mental resu
placement
ces.
d data acquis
r column sp
quired instr
The instrum
ed configura
ge number
re, from the
nal simpler
entation lay
eight levels
y selected in
rumentation
he lateral d
aluate the o
n horizonta
an externa
heights, na
Ts are fixed
o the colum
ed during th
the referre
ver, this eff
ults, by corr
component
ition systems
pecimens are
rumentation
mentation ha
ation that w
of displacem
e first tests
scheme desc
out which a
s, plus the
n order estim
n layout w
displacement
utofplane
al deflection
al steel fram
mely 20, 40
d approxima
mn by very
he tests, esp
ed wires int
fect was tak
ecting the h
measured
s
e typically d
n scheme fo
as started i
was not pro
ment transd
s result ana
cribed next.
allows meas
e local re
mate the av
was simplifie
t in the dire
displacemen
n along the
me to suppor
0, 90 and 1
ately 1.0m
thin steel
pecially for
troduced a
ken into ac
horizontal L
by the ve
49
due to
or the
in the
operly
ducers
alysis,
.
suring
elative
verage
ed by
ection
nt.
e two
rt the
140cm
away
wires.
large
small
count
LVDT
ertical
50 Chapter 3
a)
b)
Figure 27  Displacement instrumentation scheme adopted: a) lateral displacement
transducers for the EastWest direction; b) lateral displacement transducers for the
NorthSouth direction lateral displacement and relative displacement transducers
along the East and West face
In the columns West and East faces, LVDTs were fixed at different heights (but
at the same level of the horizontal LVDTs), in order to calculate the average
curvature along the column in both directions. A LVDT was introduced in the
column connection with the footing in each face of the columns to measure the
crack that is expected to open in the base of the column. In fact, considering the
test setup, the maximum moment will be occur at the column base and, therefore,
the crack likely to open in the column base can be associated with the yielding
penetration of the vertical reinforcing steel bars.
The vertical transducers used for measuring the relative displacement along the
column height were fixed with inverted rollers (Figure 29) in order to accommodate
two orthogonal direction motions.
Finally, although the column foundation is considered well fixed to the laboratory
reaction floor, spurious footing movements were monitored with LVDTs and the
axial load of the prestressing bars was also recorded through strain gauges readings
Test Program
Figure
Figure 29
to measur
m
e 28 Extern
d
a)
Inverted ro
re the relative
nal steel fram
displacements
ollers to supp
e displacemen
me to support
s in the two o
port vertical t
nts along the
view; b) deta
the LVDTs
orthogonal d
b
transducers i
e column heig
ail view
to measure t
irections
b)
n the West f
ght in biaxial
the horizonta
face of the co
l tests: a) Ge
51
al
olumn
neric
Chapter 4
Analysis of Experimental Test
Results
4.1 General overview
In the previous chapter the test program was introduced, namely the
characteristics of the specimens, the imposed load paths, the mechanical properties
of the materials and the test setup. This chapter presents the results of the overall
testing campaign, resulting from the visual observations and from the measured
parameters.
Based on the visual observation and on the test data, the damage evolution is
analysed for each single specimen and for each group of columns. Based on the
sheardrift curves, global parameters of the response are analysed, namely the
ultimate ductility, stiffness and strength degradation, as well as the coupling effect
due to the biaxial loading. In all these analyses, the differences between the
uniaxial and biaxial test results are compared. The evolution of the dissipated
energy during each test is analysed, and expressions for the equivalent viscous
damping were derived. Finally, an expression to evaluate the damage index of RC
elements under biaxial loading was proposed.
The test results and its analyses were submitted for publication in peer reviewed
international journals. A first paper where the test results are presented for the
first tested columns was accepted for publication in the Journal of Earthquake and
Tsunami [37]. A comparative analysis of the global response parameters of the
columns tested was included in a paper accepted for publication in the Earthquake
Engineering and Structural Dynamics [123]. The work relative to the analysis of
54 Chapter 4
the columns response in terms of damage and the proposal of the damage index
expression was reported in a paper submitted to the Bulletin of Earthquake
Engineering [124]. A discussion on the energy dissipation evolution and equivalent
viscous damping analysis for each column tested was published in the Engineering
Structures [125].
4.2 Damage evolution in the tested columns
4.2.1 Observed damage
It is generally accepted that, when RC elements are subjected to biaxial horizontal
loading, the stiffness and strength deterioration is intensified, and consequently the
damage evolution of the element is increased or occurs for earlier demand
levels [32, 48] when compared with uniaxial loading conditions.
Aiming at a detailed observation of the damage evolution during the cyclic loading
within the present work testing campaign, each test was stopped at the end of the
last cycle of each displacement level in order to highlight and register new cracks in
the last cycle and/or the evolution of existing ones. Visual observation of the
damage evolution during the tests yielded the information described in the
following paragraphs.
For all tests, a horizontal crack was observed at the base of the column (see Figure
30), associated with the maximum bending moment in this region. This crack is
not associated with a construction joint, since all columns were cast with the
respective footing in one single phase. Tsuno and Park [66] have also reported this
type of crack in an experimental testing campaign. It is associated with the yield
penetration of the vertical reinforcing steel bars. The bottom end of these bars is
well anchored at the footing block, with 90 reinforcement bents, so that no pull
out of the bars was observed.
In all the tests (for uniaxial or biaxial loading), horizontal cracks distributed along
the column length (associated with the flexural dominant columns response) were
observed for early demand stages as depicted in Figure 31, which illustrates the
crack evolution for test PB01N09.
For each drift demand level, the biaxial loading induces a higher level of damage in
the column base than the uniaxial loading. In the biaxial tests, after horizontal
cracks form, larger lateral demands initiate concrete spalling in the column corners
(see Figure 32b). For uniaxial tests, when concrete spalling is observed in the
column corners, it promptly expands along the whole section width of the column
(see Figure 32c). For biaxial tests, a bar located at the corner of the column base is
always the first to break (see Figure 32d).
Analysis of E
In all cas
observed (
Figure
Figure 34
30cm cros
to an axia
damage ob
axial load.
Experimental T
es where th
(see Figure 3
e 30 Exampl
Figure 31
illustrates t
ssection an
al load of 21
bserved for
.
=5m
Test Results
he plastic h
33).
le of the crac
Crack evolu
the damage
nd loaded w
10kN (=0.1
each drift d
=10mm
hinge region
ck in the base
ution in the E
evolution in
with a quad
1) and N24
demand was
=1
n comprised
e of the colum
East face of c
n columns N
rangular loa
to an axial
s higher for
5mm
d a stirrup
mn (specimen
column PB01
N15 and N24
ad path. N1
load of 650
the column
=20mm
p, its failure
n PB01N09)
1N09
4, both with
15 was subj
0kN (=0.2)
n with the h
=30
55
e was
h 30 x
jected
. The
higher
mm
56
Figur
ov
re 32 Dama
verall column
Figure
a)
c)
age states: a)
n section wid
e 33 Stirrup
) cracking; b)
th and longit
r
p failure in co
) spalling at
tudinal reinfo
rupture
olumn PB12
b)
d)
the column c
orcement buc
N11 due to h
corner; c) spa
ckling; d) cor
hoop rupture
Chapter 4
alling at the
rner bars
Analysis of E
P
B
1
2

N
1
5
N
=
2
1
0
k
N
(
=
0
.
1
)
P
B
1
2

N
2
4
N
=
6
5
0
k
N
(
=
0
.
2
)
drift
Figure 34
4.2.2 D
Typically,
gravity loa
These regi
curvatures
namely cr
and reinfo
expression
exhaustive
Bayrak [12
It is difficu
tests. How
and the sev
damaged z
Table 7 (s
As expecte
proportion
the length
strong dire
is a functi
tested wit
length is v
Experimental T
0.33%
(5mm)
4 Damage e
load
amage d
slender R
ads, present
ions are com
s [126] are o
osssection
orcement st
ns have been
e review of
27].
ult to evaluat
wever, some c
verely damag
zone in each
ee Figure 36
ed, in the u
nal to the cr
h of the dam
ection. In re
on of the la
th the same
very similar
Test Results
1.00%
(15mm
evolution in t
d ratios but t
distribut
RC columns
t severe dam
mmonly refe
observed. T
dimensions,
eel) and th
n proposed
these form
te the plastic
corresponden
ged column zo
tested colum
6 for examp
uniaxial tests
rosssection
mage region
ectangular c
arger crosss
e axial load
for uniaxial
%
m)
1
(2
the columns
to the same h
tion
, subjected
mage in the
erred to as p
The plastic h
, level of ax
he confinem
for the esti
mulations ca
c hinge length
nce can be est
ones (see Fig
mn was measu
in
ples).
s the length
height. For
n was highe
columns bia
section dime
d ratio (colu
l and biaxia
1.67%
25mm)
N15 and N24
horizontal loa
to earthq
e regions wi
plastic hinge
hinge length
xial load, m
ment level in
imation of t
an be foun
h based on th
tablished bet
gure 35). For
ured and the
h of the dam
r rectangula
er when th
axially teste
ension. For
umns N13 t
al tests.
2.00%
(30mm)
4, subjected t
ading path
uake loads
th larger be
es where the
h depends o
material pro
n the hinge
the plastic h
d in the w
he instrument
ween the pla
this purpose
recorded val
maged zone
ar columns u
e column w
d, the dama
square colum
to N16), th
3.00
(45m
to different a
and suppo
ending mom
e largest ine
on many fa
operties (con
e region. Va
hinge length
work of Bae
tation used i
astic hinge len
e, the length
lues are prese
is approxim
uniaxially t
was loaded
aged zone l
mns (30 x 3
he damaged
57
0%
mm)
axial
orting
ments.
elastic
actors,
ncrete
arious
h. An
e and
n the
ngth
of the
ented
mately
ested,
in its
ength
30cm)
zone
58
Comp
load
loadi
Tsun
bidir
zone
lengt
lengt
Fi
paring the
path is see
ing paths.
no and Park
ectional loa
length in t
th for rectan
th estimated
igure 36 Ex
different bi
en to induce
k [66] stated
ding. Simila
the columns
ngular colum
d for the col
Figure 35
xamples of da
differe
iaxial loadin
e a damage
d that the
ar conclusion
s tested in
mns under b
umns stron
Plastic hin
amage distrib
ent uniaxial a
ng paths for
ed zone leng
plastic hing
ns can be d
this work.
biaxial loadi
ng direction.
nge length (ad
bution in a 30
and biaxial lo
r all tested
gth slightly
ge zone leng
rawn from t
Therefore, t
ng condition
.
dapted from
0 x 50cm RC
oading paths
columns, t
larger than
gth is not a
the analysis
the plastic
ns can be ta
[36])
C column subj
Chapter 4
the circular
n the other
affected by
s of damage
hinge zone
aken as the
jected to
Analysis of Experimental Test Results 59
Table 7 Length of the damage zone
Series Column
Cross
section
[cm x cm]
Damage zone
[cm]
Displacement
path type
Minimum Maximum Average
1
PB01N01
20 x 40 0.04
13 16 14.5 1
PB02N02 2
PB12N03 12 18 15 3
PB12N04 13 16 15 4
2
PB01N05
30 x 40 0.12
14 17 15.3 1
PB02N06 16 29 22.3 2
PB12N07 14 23 19.5 4
PB12N08 14 30 21.25 5
PB12N17 24 28 30.4 6
3
PB01N09
30 x 50 0.08
26 31 29 1
PB02N10 13 19 16 2
PB12N11 24 30 27 4
PB12N12 20 23 21 5
PB12N18 23 35 28 6
4
PB01N13
30 x 30 0.1
11 14 13 1
PB12N14 10 15 12.5 4
PB12N15 12 14 13 5
PB12N16 13 16 15.5 6
5
PB12N19
30 x 50
0.045 24 35 30 4
PB12N20 0.09 17 27 22 4
6
PB12N21
30 x 40 0.12
24 30 27 4
PB12N22 30 32 31 5
7
PB12N23
30 x 30 0.2
38 42 41 4
PB12N24 32 36 34 5
4.2.3 Damage versus drift demand
For each tested column, it was identified the drift corresponding to the
characteristic damage stages, namely the onset of cracking, the concrete cover
spalling, the longitudinal reinforcement buckling and the first fracture of the
longitudinal bars (see examples in Figure 32). To understand the influence of
biaxial loading in the damage evolution, comparison was made of the drift values
corresponding to each damage type for all tested columns. The conventional failure
of the columns for a strength reduction of 20% relative to the maximum strength,
as adopted in Park and Ang [128], is also compared.
The drift corresponding to the onset of damage for each of the different load paths
(uniaxial and biaxial) is compared in Figure 37. As expected, each type of damage
(cracking, spalling, etc.) occurs earlier for biaxial load paths than for uniaxial tests,
i.e., for lower drift demands (see Figure 37). By comparing the drift demand
60 Chapter 4
corresponding to each damage type for biaxial and uniaxial tests, in the columns
strong direction, the following can be stated:
 Concrete cracking under biaxial load occurs for a drift demand of about
60% of the corresponding load in a uniaxial test.
 Concrete spalling in biaxial tests occurs for a drift in the range of 5075%
of that observed in corresponding uniaxial tests.
 In biaxial tests, bar buckling occurs for a drift between 65% and 75% of
that observed in uniaxial tests.
 In all cases, conventional rupture occurs for drift demands close to the
corresponding drift values when reinforcement steel bars buckling is
observed.
 The drift ratios (biaxial relative to the uniaxial test results) for spalling and
buckling are compared for the columns strong direction in Figure 38. A
good correlation in terms of drift reduction is observed for the damage
states due to biaxial and uniaxial loading.
 The influence of the axial load and of axial load ratio on the observed
damage is illustrated in Figure 39, while the effects of cycles repetition on
the damage evolution are represented in Figure 40.
 As shown in Figure 39, higher axial forces anticipate all damage states,
being more pronounced for buckling and conventional rupture. Also, higher
levels of axial load tend to reduce the drift difference between two
consecutive damage states. After spalling, the subsequent damage states
occur for practically the same drift demand.
 The nonrepetition of the cycles for each displacement level, implies a larger
drift demand to reach each damage state (see Figure 40), confirming the
influence of cyclic load repetition in the anticipation of each type of
damage.
Analysis of Experimental Test Results 61
a) b)
c)
d)
Figure 37 Drift associated with each damage state for different load paths (uniaxial and
biaxial)
Figure 38 Correlation between the ratios (biaxial/uniaxial) of drift demand corresponding
to concrete spalling and bar buckling
0 1 2 3 4 5
X
Y
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01M01
PB02N02
PB12N03
PB12N04
Cracking
Spalling
Buckling
Conv. Rupture
1st bar fracture
0 1 2 3 4 5
X
Y
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N05
PB02N06
PB12N07
PB12N08
PB12N17
Cracking
Spalling
Buckling
Conv. Rupture
1st bar fracture
0 1 2 3 4 5
X
Y
1st bar fracture
Conv. Rupture
Buckling
Spalling
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N09
PB02N10
PB12N11
PB12N12
PB12N18
Cracking
0 1 2 3 4 5
X
Y
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N13
PB02N14
PB12N15
PB12N16
Cracking
Spalling
Buckling
Conv. Rupture
1st bar fracture
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
Experimental data
Best fitted curve (R
2
=0.998)
B
i
a
x
i
a
l
B
u
c
k
l
i
n
g
D
i
s
p
l
a
c
e
m
e
n
t
A
g
a
i
n
s
t
U
n
i
a
x
i
a
l
B
u
c
k
l
i
n
g
D
i
s
p
l
a
c
e
m
e
n
t
(
%
)
Biaxial Spalling Displacement Against
Uniaxial Spalling Displacement (%)
y = 1.04721x
0.50 0.55 0.60 0.65 0.70 0.75 0.80
0.55
0.60
0.65
0.70
0.75
0.80
62 Chapter 4
a)
b)
Figure 39 Influence of the axial load on the drift associated with each damage state for
biaxially loaded columns
Figure 40 Influence of the cycle repetition on the drift associated with each damage state
for biaxially loaded columns
4.2.4 Performance objective
Performance objectives are statements of acceptable performance of a given
structure. The performance target can be specified by limits of any response
parameter, such as stresses, strains, displacements, accelerations, etc. [129].
Considering the definition of the performance levels, various international
guidelines were considered in this study, namely SEAOC VISION 2000 [130], ATC
40 [88], FEMA 273, FEMA 274 and FEMA 356 [90, 131]. These documents specify
the allowable drifts for design and assessment of structures that are expected to
experience dynamic loading. In general, the allowable drifts depend on the
structural material and the structural system. Table 8 summarizes the
representative damage in columns for RC frame buildings for each performance
level (according to FEMA 356 [90]), while the drift limits for each structural
performance level, according to FEMA 356 and VISION 2000 proposals, are
presented in Table 9.
0 1 2 3 4 5
X
Y
1st bar fracture
Conv. Rupture
Buckling
Spalling
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N11 (N= 300; u=0.080)
PB02N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
Cracking
0 1 2 3 4 5
X
Y
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N14 (N= 210; u=0.1)
PB02N15 (N= 210; u=0.1)
PB12N23 (N= 650; u=0.2)
PB12N24 (N= 650; u=0.2)
Cracking
Spalling
Buckling
Conv. Rupture
1st bar fracture
0 1 2 3 4 5
X
Y
O
b
s
e
r
v
e
d
d
a
m
a
g
e
Drift (%)
PB01N07 (N= 300; u=0.12)
PB02N08 (N= 300; u=0.12)
PB12N21 (N= 620; u=0.12)
PB12N22 (N= 620; u=0.12)
Cracking
Spalling
Buckling
Conv. Rupture
1st bar fracture
Analysis of Experimental Test Results 63
Table 8 Structural Performance Levels and Damage for RC columns adapted from FEMA
356 [90]
Structural Performance Levels
Immediate Occupancy Life Safety Collapse Prevention
Minor hairline cracking.
Limited yielding possible at
a few locations. No crushing
Spalling of cover
and shear cracking
for ductile columns
Extensive cracking and
hinge formation in
ductile elements
Table 9 Drift limits according to the (a) FEMA 356 [90], (b) VISION 2000 [130]
Performance Level
Immediate
Occupancy
Damage
Control
Life
Safety
Structural Stability
(a) Drift limits according to the FEMA 356 [90]
Drift
Limit
1% 12% 2% 4%
Fully Operational Operational
Life
Safety
Near Collapse
(b) Drift limits according to the VISION 2000 [130]
Drift
Limit
0.2% 0.5% 1.5% 2.5%
For the tested columns, the drift range for each damage state reported in the
previous section is summarised in Table 10. In Figure 41, the experimental drift
ranges are compared with the drift limits provided in FEMA 356 [90] and
VISION 2000 [130] for each performance level. From the analysis of Figure 41 and
Table 10, it is observed that, even considering the reduction of the drift for each
damage state associated with the biaxial loading,
Comparing the structural performance levels and damage for RC columns (Table
8), adapted from FEMA 356 [90], with the observed drift ranges (Table 10) for
each damage state, it was observed that: in the drift range corresponding to
Immediate Occupancy limit state (0 to 1%) it was observed mainly concrete
cracking; for the Life Safety limit state (1 to 2%) it was observed mainly concrete
spalling and in some biaxial tests bar buckling; in the Near Collapse and Collapse
Prevention (in the drift range between 2 and 4%), it was observed bar bucking and
the conventional rupture. As so, the drift limits considered by those international
guidelines and the damage observed experimentally is in good agreement.
64 Chapter 4
Table 10 Drift range obtained from experimental tests
Damage state
Drift range (%)
uniaxial tests biaxial tests
Cracking 0.250.55 0.100.55
Spalling 2.303.90 1.002.60
Bar Bucking 2.654.65 1.403.60
Conventional rupture 3.005.00 1.404.30
Figure 41 experimental drifts for each damage state and drift limits according to the
FEMA 356 [90], VISION 2000 [130]
4.3 Forcedisplacement hysteretic behaviour
4.3.1 Global analysis
In order to study the influence of the biaxial loads on the behaviour of RC
columns, the measured displacement and shear force paths (along the X and Y
directions) were analysed. In the tests, no significant torsional rotation of the
columns was observed. Due to the large number of tests performed, only a few
examples of the results are presented in Figure 42 to 44, but the discussion refers to
the complete testing program results which are fully detailed in Appendix A.
0 1 2 3 4 5
Cracking
Biaxial test
Uniaxial test
Legend
VISION 2000 / FEMA 356
Performance Level
Conventional rupture
Bar buckling
Concrete spalling
V
I
S
I
O
N
2
0
0
0
V
I
S
I
O
N
2
0
0
0
F
E
M
A
3
5
6
I
n
t
e
r
n
a
t
i
o
n
a
l
g
u
i
d
e
l
i
n
e
s
Drift (%)
E
x
p
e
r
i
m
e
n
t
a
l
R
e
s
u
l
t
s
Collapse
Near Collapse/
Structural Stability
Life Safety
Operational/
Damage Control
Fully Operational/
Immediate Occupancy
Legend
Analysis of Experimental Test Results 65
Figure 42 Global results of rectangular column PB12N07 for rhombus load path
Figure 43 Global results of rectangular column PB12N12 for rectangular load path
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
12010080 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
60 40 20 0 20 40 60
60
40
20
0
20
40
60
S
h
e
a
r
X
(
k
N
)
Drift X (%)
1412 10 8 6 4 2 0 2 4 6 8 10 12 14
Displacement ductility factor
Displacement ductility factor
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
141210 8 6 4 2 0 2 4 6 8 10 12 14
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
7
1
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
20
16 12 8
4
18
14
10
6
2
19
15
11
11
7
3
17
13
9 5
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
20
X
Y
PB12N07
3 2 1 0 1 2 3
180
150
120
90
60
30
0
30
60
90
120
150
180
3 2 1 0 1 2 3
180
150
120
90
60
30
0
30
60
90
120
150
180
180150120 90 60 30 0 30 60 90 120 150 180
180
150
120
90
60
30
0
30
60
90
120
150
180
50 40 30 20 10 0 10 20 30 40 50
50
40
30
20
10
0
10
20
30
40
50
20
19
15
14
10
9
5
4
18
17
13
12
8
7
3
2
Displacement ductility factor
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
5 4 3 2 1 0 1 2 3 4 5
19
18
14
13
9
8
4
3
12
7
2
17
16
11
6
Displacement ductility factor
S
h
e
a
r
X
(
k
N
)
Drift X (%)
10 8 6 4 2 0 2 4 6 8 10
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3 2
1
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
1
20 19
18
16
17
15 14
13 12
11
10
9
8 7
6
5 4
3 2
1
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N12
d)
c)
b) a)
d) c)
b)
a)
66 Chapter 4
Figure 44 Global results of rectangular column PB18N12 for circular load path
Aiming at evaluating the effect of the biaxial load paths on the stiffness and
strength degradation of the RC columns, the maximum envelopes of the sheardrift
hysteresis curves and the lateral peaktopeak stiffness degradation were analysed.
Additionally, the N11, N19 and N20 tests were used to analyse the influence of the
total axial load and the axial load ratio on the initial stiffness and the maximum
strength. Therefore, the hysteresis curves envelopes for the tested columns are
plotted in Figure 45 to 48, and a summary of the main findings is presented in
Table 11 to Table 16. Because of the loading paths adopted in the experiments, the
first cycle of each displacement level in the biaxial tests always occurs in the same
direction (i.e., the positive X direction). This induces a different response in the
first cycle of each displacement amplitude level, but the effect is reduced in the
subsequent cycles for the same displacement amplitude.
The analysis of the forcedisplacement hysteretic behaviour focused on the
following main issues: i) the comparison and identification of the main differences
in the shapes of the envelopes of the uniaxial and biaxial response test results; ii)
the evaluation of the lower ductility in biaxial tests when compared with the
corresponding uniaxial tests; iii) the interpretation of the strong coupling between
the responses in the two directions as observed in the 2D tests; and iv) the
characterisation of the correlation found between the maximum strength and the
yield force.
40 20 0 20 40
40
20
0
20
40
200 150 100 50 0 50 100 150 200
100
50
0
50
100
3 2 1 0 1 2 3
200
150
100
50
0
50
100
150
200
3 2 1 0 1 2 3
100
50
0
50
100
4
3
2
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
PB12N18
X
Y
1
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
4
3
2
1
4
3
2
1
4
3
2
1
Displacement ductility factor
S
h
e
a
r
X
(
k
N
)
Drift X (%)
5
7
8
6
11
12
10
9
14
16
15
13
4 3 2 1 0 1 2 3 4
6
Displacement ductility factor
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
6
7
8
5
10
12
11
9
14
15
16
13
6
7
8
5
10
11
12
9
14
15
16
13
6
7
8
5
10
11
12
9
14
15
16
13
7 6 5 4 3 2 1 0 1 2 3 4 5 6 7
d) c)
b)
a)
Analysis of Experimental Test Results 67
The results analysis can be summarised as described in the following:
 In the sheardrift curves envelopes, four main stages can be identified in
the, corresponding to: i) the precracking response; ii) the postcracking
behaviour, until the reinforcement steel begins to yield; iii) a moreorless
evident plateau in the postyielding hardening zone and iv) a softening
phase. These four stages are clear in both the uniaxial and the biaxial tests.
However, in the biaxial tests, the plateau tends to be shorter and the
softening is more pronounced, i.e., a more abrupt decay of the column
strength is observed with increasing lateral deformation demands.
 The initial column stiffness is not significantly affected by the biaxial
loading path (Figure 45) in either direction.
 Columns having the same geometry and similar axial load ratios but made
with different concrete grades [N11 (=0.08 and N=300kN) and N19
(=0.09 and N=600kN)] showed similar initial stiffness values but a larger
maximum strength was observed for the column (N19) with the higher axial
load (see Figure 46).
 However, columns having the same geometry and the same total axial load,
i.e., N=300kN for both N11 (=0.08) and N20 (=0.045), though made
with different concrete grades, exhibited similar maximum strength values;
however, larger stiffness was obtained for the column N11 having the higher
axial load ratio (Figure 46 and Figure 47).
 For each cycle of the biaxial test on the square columns, the maximum
horizontal force in the Y direction was between 10 and 20% smaller than
that in the X direction (Figure 48 and Table 14), an effect that has been
also reported by other authors [66, 132]. As mentioned before, this effect is
induced by the first push along the loading path, which always occurs in
the X direction.
 As expected, when the maximum columns strength of the in each direction
for each biaxial test was compared with the results from the corresponding
uniaxial test, lower values were found in the biaxial cases. Furthermore,
this effect is more pronounced in the square columns.
 The biaxial loading produced a reduction of approximately 20% in the
maximum strength of the rectangular columns in their weak direction (Y),
while reductions ranging from 25 to 30% were observed in the square
columns. The strength reduction in the weaker direction is always larger
than that observed in the corresponding test along the stronger direction
(X), which is approximately 8% in the rectangular columns and 15% in the
68 Chapter 4
square columns. This is because, for rectangular columns, the response in
the strong direction is less affected by the damage previously produced in
the weak direction than does the response in the weak direction due to
damage in the strong direction.
Figure 45 Envelopes for different load paths
5 4 3 2 1 0 1 2 3 4 5
80
60
40
20
0
20
40
60
80
S
h
e
a
r
(
k
N
)
Drift X (%)
PB01N01
PB12N03
PB12N04
5 4 3 2 1 0 1 2 3 4 5
80
60
40
20
0
20
40
60
80
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB02N02
PB12N03
PB12N04
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
S
h
e
a
r
(
k
N
)
Drift X (%)
PB01N05
PB12N07
PB12N08
PB12N17
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB02N06
PB12N07
PB12N08
PB12N17
5 4 3 2 1 0 1 2 3 4 5
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
S
h
e
a
r
(
k
N
)
Drift X (%)
PB01N09
PB12N11
PB12N12
PB12N18
5 4 3 2 1 0 1 2 3 4 5
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB02N10
PB12N11
PB12N12
PB12N18
5 4 3 2 1 0 1 2 3 4 5
100
80
60
40
20
0
20
40
60
80
100
S
h
e
a
r
(
k
N
)
Drift X (%)
PB01N13
PB12N14
PB12N15
PB12N16
5 4 3 2 1 0 1 2 3 4 5
100
80
60
40
20
0
20
40
60
80
100
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB01N13
PB12N14
PB12N15
PB12N16
a) b)
c) d)
e) f)
g) h)
Analysis of Experimental Test Results 69
Figure 46 Envelopes for different levels of axial force
Figure 47 Envelopes for same load path with and without cyclic repetition
Figure 48 Effect of the biaxial load path in the maximum columns strength
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
(
k
N
)
Drift X (%)
PB12N11 (N= 300; u=0.080)
PB12N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB12N11 (N= 300; u=0.080)
PB12N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
X
Y
S
h
e
a
r
(
k
N
)
Drift X (%)
PB12N14 (N= 210; u=0.1)
PB12N15 (N= 210; u=0.1)
PB12N23 (N= 650; u=0.2)
PB12N24 (N= 650; u=0.2)
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
X
Y
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB12N14 (N= 210; u=0.1)
PB12N15 (N= 210; u=0.1)
PB12N23 (N= 650; u=0.2)
PB12N24 (N= 650; u=0.2)
5 4 3 2 1 0 1 2 3 4 5
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
X
Y
S
h
e
a
r
(
k
N
)
Drift X (%)
PB01N07 (N= 300; u=0.08)
PB12N08 (N= 300; u=0.08)
PB12N21 (N= 620; u=0.08)
PB12N22 (N= 620; u=0.08)
5 4 3 2 1 0 1 2 3 4 5
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
X
Y
S
h
e
a
r
(
k
N
)
Drift Y (%)
PB01N07 (N= 300; u=0.08)
PB12N08 (N= 300; u=0.08)
PB12N21 (N= 620; u=0.08)
PB12N22 (N= 620; u=0.08)
N
o
n

d
i
m
e
n
s
i
o
n
a
l
M
a
x
i
m
u
m
H
o
r
i
z
o
n
t
a
l
F
o
r
c
e
i
n
B
i
a
x
i
a
l
T
e
s
t
A
g
a
i
n
s
t
U
n
i
a
x
i
a
l
T
e
s
t
x+ x y+ y x+ x y+ y x+ x y+ y
0.0
0.2
0.4
0.6
0.8
1.0
Circular Quadrangular Rombus
Horizontal Displacement path
30x50 30x30
a) b)
c) d)
70 Chapter 4
Table 11 Summary of test results for columns N01 to N04
Specimen
Loading
direction
F
max
[kN]
y
(mm)
F
y
[kN]
=
u
/
y
PB01N01 X
+ 68.4
5.83
53.0
8.6
 73.0 51.5
PB02N2 Y
+ 35.3
5.39
24.89
 34.4 24.6
PB12N03
X
+ 69.4
6.01
50.6
7.5
 68.7 46.9
Y
+ 33.3
5.72
19.5
11.4
 31.4 16.9
PB12N04
X
+ 69.7
5.29
52.41
9.1
 64.5 42.57
Y
+ 24.7
4.23
19.5
11.4
 27.3 17.4
Table 12 Summary of test results for columns N05 to N08 and N17
Specimen
Loading
direction
F
max
[kN]
y
(mm)
F
y
[kN]
=
u
/
y
PB01
N05
X
+ 102.5
4.70
74.3
9.6
 115.3 76.9
PB02N6 Y
+ 91.5
3.78
55.8
18.5
 83.4 60.9
PB12
N07
X
+ 112.1
5.11
80.3
6.8
 114.5 81.5
Y
+ 71.1
5.18
51.6
6.7
 73.9 58.5
PB12
N08
X
+ 98.2
4.27
64.8
5.9
 102.6 73.3
Y
+ 66.4
4.89
44.7
5.1
 74.4 49.9
PB12
N17
X
+ 131.4
6.96
97.0
3.71
 104.8 86.6
Y
+ 80.7
4.75
52.4
5.43
 77.1 49.2
Analysis of Experimental Test Results 71
Table 13 Summary of test results for columns N09 to N12 and N18
Specimen
Loading
direction
F
max
[kN]
y
(mm)
F
y
[kN]
=
u
/
y
PB01N09 X
+ 167.9
5.77
115.7
7.8
 164.6 113.6
PB02N10 Y
+ 96.7
10.75
70.7
7.0
 102.7 79.1
PB12N11
X
+ 173.7
5.04
121.8
6.9
 163.5 110.3
Y
+ 69.8
5.49
50.2
6.4
 82.8 61.6
PB12N12
X
+ 153.4
4.74
112.0
5.3
 156.7 109.5
Y
+ 74.6
8.87
51.6
2.8
 88.1 69.0
PB12N18
X
+ 175.3
6.63
116.1
4.63
 155.7 113.9
Y
+ 80.1
4.58
52.2
6.7
 83.6 50.8
Table 14 Summary of test results for columns N13 to N16
Specimen
Loading
direction
F
max
[kN]
y
(mm)
F
y
[kN]
=
u
/
y
PB01N13 X
+ 71.4
12.14
61.8
4.9
 73.8 56.2
PB02N14
X
+ 62.1
10.12
47.5
4.4
 66.8 50.6
Y
+ 54.0
10.33
45.5
4.4
 59.8 46.1
PB12N15
X
+ 58.1
11.11
43.9
3.1
 65.1 46.4
Y
+ 52.8
9.09
45.0
3.8
 53.7 47.1
PB12N16
X
+ 59.8
12.47
51.2
4.0
 58.8 43.8
Y
+ 47.7
10.69
34.4
4.7
 49.8 37.8
72 Chapter 4
Table 15 Summary of test results for columns N19 to N24
Specimen
Loading
direction
F
max
[kN]
y
(mm)
F
y
[kN]
=
u
/
y
PB12N19
X
+ 206.67
10.37
176.85
3.9
 208.94 152.64
Y
+ 97.83
8.58
88.62
4.6
 115.02 89.64
PB12N20
X
+ 166.41
5.53
110.67
7.4
 156.02 95.02
Y
+ 73.50
4.99
55.42
10.3
 72.67 53.73
PB12N21
X
+ 157.38
5.30
98.27
12.3
 148.49 92.61
Y
+ 96.33
5.01
65.57
11.2
 107.85 69.09
PB12N22
X
+ 150.48
4.44
101.83
10.0
 136.73 84.72
Y
+ 93.12
3.75
55.94
11.4
 106.83 56.94
PB12N23
X
+ 103.27
3.53
66.25
6.0
 99.91 58.79
Y
+ 96.74
3.17
48.60
6.6
 97.39 55.76
PB12N24
X
+ 104.31
3.96
72.94
6.5
 97.04 59.22
Y
+ 77.27
3.78
54.16
6.5
 102.07 59.79
4.3.2 Evaluation of the yielding displacement
The yield displacement is fundamental for differentiating the response of a RC
element in the elastic and postelastic regimes [133]. Recently, several methods
have been proposed for estimating the yield curvature or yield displacement of an
RC section or element [133136], and there are significant differences between such
proposals. In the present work, the yield displacement was derived from tests on
RC columns to characterise and compare the columns response in terms of the
displacement ductility.
On the basis of analytical results, Priestley (mm)[134] concluded that for RC
members, the yield curvature is essentially independent of the reinforcement
content and the axial load level, but it is clearly dependent on the yield strain and
the section depth.
Analysis of Experimental Test Results 73
Similarly, from experimental results on RC elements, other authors define the yield
displacement as the deflection of an equivalent perfect elastoplastic system with
energy absorption equal to that observed in the real system [137].
Park [136] proposed a procedure for the cyclic testing of RC columns based on
previous assessments of the yield displacement. Essentially, the test is run in two
stages, where the first stage corresponds to the evaluation of the yield
displacement. First, a lateral load is applied monotonically to the specimen in one
direction until the load is either equal to the first calculated or measured yield or
equal to threequarters of the calculated ideal strength of the specimen, whichever
is less. Then, the specimen is loaded in the opposite direction using the same
procedure. The average of the displacements obtained in the two directions,
corresponding to the intersection points of the secant stiffness and the horizontal
lines for the ideal strength of the specimen, defines the reference yield displacement
[66].
The procedure adopted in this work for the determination of the yield displacement
is based on the method proposed by Park [136], but it was adapted to be used
along with the complete cyclic test results, whilst considering the possible
hardening of the postyield response. The adopted procedure is described as follows
and illustrated in Figure 49:
 Evaluation of the maximum strength of the specimen in both test
directions;
 Identification of the cycle in which the strength is lower than threequarters
of the previously evaluated maximum strength of the specimen;
 Calculation of the secant stiffness (K
y
) for the cycle identified in step 2;
 Adjustment of the branch corresponding to the postyield stiffness (K
pl
);
 Determination of the intersection points of K
y
and K
pl
for each direction,
giving the yield displacement in each direction. The average of the yield
displacements obtained in the two directions (d
y
+
and d
y

) provides the
reference yield displacement (
y
).
74 Chapter 4
Figure 49 Proposed method for the definition of yielding displacement
The method adopted to evaluate the yield displacement in this analysis has a clear
advantage over the estimation by simplified formulae because the method is based
on experimental test results. Moreover, procedures that define the yield point based
on an equivalent elastoplastic system with equal energy absorption may not
properly represent the influence of the initial stiffness and the postyield stiffness
on the yield displacement. In contrast, the procedure adopted here, inspired by the
proposals of Park [136], defines the yield displacement in a prior testing phase,
allowing for the characterisation of the subsequent cyclic load path.
4.3.3 Influence of the biaxial loading path on the
columns ultimate ductility
From the analysis of the maximum ductility values presented in Table 11 through
Table 15 and from the sheardrift envelope curves, it is apparent that biaxial
loading results in lower ultimate column ductility than does uniaxial loading. The
following conclusions can be drawn from the results:
 Amongst the imposed biaxial loading paths, the quadrangular path
produces the lowest ultimate ductility.
 In the rectangular columns, loading in the weak direction (Y) yields
ductility reductions (of approximately 50 to 75%) relative to those in the
corresponding uniaxial case that are greater than those in the strong
direction (approximately 35%). This is due to the damage state
anticipation, particularly when bar buckling occurs, which causes the
ultimate point to be reached earlier in the weak direction than in the strong
direction. It is worth mentioning that a typical variation of 20% was
observed in the yielding displacement, relative to the uniaxial case, but
d

y
d
+
y
3/4 of maximum strength
3/4 of maximum strength
maximum strength
K
y
K

pl
S
h
e
a
r
Drift
K
+
pl
maximum strength
cycle with strength inferior
to three quarters of the
evaluated maximum strength
Analysis of Experimental Test Results 75
without a systematic increasing or decreasing trend. Although the variation
appears more pronounced in the weak direction, there is little effect on the
computation of the ultimate ductility, in contrast with the influence of the
ultimate point variation.
 In the square columns, the ductility capacity reduction (with the biaxial
loading path) was generally between 10 and 20% for both directions,
although the reduction reached 35% in one case.
4.3.4 Coupling effect in the columns biaxial response
From the analysis of the hysteretic response of the tested columns, the following
results were found:
 The unloading stiffness of the uniaxial load path is higher than that
observed in the biaxial tests.
 Significant differences were found between the reloading stiffness values
obtained from the uniaxial and the biaxial tests. The reloading stiffness of
the quadrangular load path is larger than that observed in the uniaxial
tests.
 A clear pinching effect was observed in the rhombus loading path,
especially in the columns weak direction.
 In all cases with circular loading paths, the sheardrift hysteresis loops
become wider and more rounded at the peaks as the drift demands increase
and damage accumulates (see, for example, the response of column N12 in
Figure 44).
The latter observation has been also reported by other authors [44, 48, 60] and is
associated with the phase lag evidenced by the angle between the vectors of the
biaxial lateral displacement and the biaxial transversal load. The phase lag leads to
a relative rotation of the biaxial loading path with respect to the biaxial
deflections. The phase lag is mainly due to the column nonlinear response, which
causes residual displacements to occur at zero force levels; this fact implies that
unloading towards zero displacement in a given direction requires imposing a force
in the sense opposite to the measured displacement. This effect is quite apparent in
the rhombus path loading shown in Figure 42 (and in the quadrangular path
loading shown in Figure 43) and it increases as the damage increases throughout
the column response.
The following comparisons can be made between the phase lag observed in the
columns tested for the circular path (Figure 50) and the results reported by
Bousias et al. [48]:
76 Chapter 4
 Bousias et al. [48] tested a square column at displacements of 20, 50, 80 and
110mm and reported the occurrence of a phase lag at each displacement
level, increasing from 10 to 40 (at displacements of 20 and 110mm,
respectively) with the increasing horizontal displacement.
 In the present test results, the phase lag also increased with the magnitude
of the horizontal displacement. However, this effect was reduced by the
larger number of cycles made and by the smooth increase of the
displacement magnitude because the imposed horizontal displacement
pattern resulted in a smooth change in the resulting damage. The large
jumps between the cycles of imposed displacement reported in the tests by
Bousias et al. [48] introduced abrupt changes in stiffness in the direction in
which the cycle started, thus magnifying the phase lag in this stage.
 In the tested columns, the phase lag began at an angle of approximately
20. In the rectangular columns in particular (N17 and N18), the phase lag
angle reached values of approximately 35 to 40 at the end of testing.
In the internal cycles, an increase in the phase lag was verified as shown in Figure
50.
Figure 50 Phase lag between measured horizontal forces and imposed horizontal
displacements for circular load paths
The phase lag effect was particularly evident in the test results for the square
columns and for both the imposed rhombus and quadrangular loading paths,
producing path rotation angle values within the range reported by other authors.
Takizawa and Aoyama [60], Otani [39] and Bousias et al. [48] observed rotation
angles of between 10 and 20 in square column tests.
Results from the tests on the rectangular columns confirm that the path rotation
depends on the section geometry. For the rhombus loading path, the rotation
observed in the tests on rectangular columns was approximately 7 to 10 in the
strong direction (X) and 30 to 35 in the weak direction (Y) (Figure 42). For the
quadrangular loading path, this effect is even more significant because the rotation
observed is within the range of approximately 3 to 6 in the strong direction and
approximately 40 to 50 in the weak direction (Figure 43).
0 1000 2000 3000 4000 5000
0
10
20
30
40
50
60
70
80
90
A
n
g
le
(
d
e
g
)
Step
PB12N16
X
Y
X
Internal cycles
0 1000 2000
0
10
20
30
40
50
60
70
80
90
X
Y
A
n
g
le
(
d
e
g
)
Step
PB12N17
X
Internal cycles
0 1000 2000
0
10
20
30
40
50
60
70
80
90
X
Y
A
n
g
le
(
d
e
g
)
Step
PB12N18
X
Internal cycles
Analysis of Experimental Test Results 77
4.3.5 Correlation between maximum strength and yield
strength
As mentioned previously, the definition of the yield force and the yield
displacement is a difficult and controversial subject. Different authors have
proposed simplified methodologies to estimate the yield strength.
The maximum and yield strengths obtained in this study are plotted in Figure 51.
Two points, one for each direction (X and Y), are presented for each biaxial test.
The bestfit line is also included in the figure. A high linear correlation is found
(with a correlation coefficient of R
2
=0.988) between the maximum strength and the
yield strength, showing that the yielding strength is approximately 70% of the
corresponding maximum strength for all of the test results. Additional correlations
were calculated to verify independently the influence of the uniaxial and biaxial
tests, the loading type (uniaxial and biaxial), the load direction (weak and strong
column directions) and the column geometry (rectangular and square) on the ratio
of the yield strength to the maximum strength. For each of the individual
comparisons made, similar values were obtained for the linear correlation factor.
Figure 51 Correlation between maximum and yielding strength
4.4 Strength degradation
From the analysis of a series of static cyclic tests of RC elements, it was
observed [138] that the strength degradation of RC columns normally increases
with cyclic loading, particularly after the maximum strength has been reached. In a
similar manner, this section aims at assessing the influence of biaxial cyclic loading
on the strength degradation of columns. Thus, for each displacement demand
0 25 50 75 100 125 150 175 200 225
0
25
50
75
100
125
150
175
200
225
y = 0.697.x
R
2
= 0.988
Uniaxial test
Biaxial test
Y
i
e
l
d
S
t
r
e
n
g
t
h
(
k
N
)
Maximum Shear Strength (kN)
78 Chapter 4
(comprising three cycles), the shear force was normalised by the maximum shear
force, which occurs during the first pull (see Figure 52). The strength degradation
analysis was based on the semicycles of negative force because the first force
incursion for each displacement amplitude was made in the positive force direction
in all of the tests.
Figure 52 Strength degradation under cyclic loading
Figure 53 shows the strength degradation results calculated from the uniaxial and
biaxial tests of columns N05 to N18 for all crosssection types, and Figure 54
demonstrates the influence of the axial force on the strength degradation.
Generally, the strength degradation exhibited consistent trends in all tested
columns, allowing the following observations to be made:
 In the uniaxial tests, a lower level of strength degradation was observed
than in the biaxial tests. This finding has also been reported by other
authors [32, 43, 48], who concluded that this property is one of the more
important aspects of the biaxial response of columns. In fact, in the biaxial
tests, each damage state (cracking, spalling, and bar buckling) starts at a
lower drift demand level than in the corresponding uniaxial test. Thus,
increased strength degradation is expected in biaxial loading conditions and
is mainly associated with concrete spalling and bar buckling, which occur at
lower demand levels.
 In uniaxial and biaxial tests, the strength degradation is practically
constant with increasing maximum ductility up to an inflexion point, which
was found to be associated with the buckling of the longitudinal
reinforcement bars.
 The loading path has a significant influence on strength degradation
(Figure 53). In all tests, both uniaxial and biaxial, the strength degradation
was limited to less than 10% until a displacement ductility factor of
approximately 3 was reached. After this point, the strength degradation
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB01N01
Analysis of Experimental Test Results 79
started increasing further, which was found more pronounced for the square
loading path.
 Figure 54 show that the axial load has a significant influence on strength
degradation. For different levels of the axial load ratio but equal axial
forces [N11 (=0.08) and N20 (=0.045), both with N=300kN], the
evolution of the strength degradation is similar. For similar values of the
axial load ratio but different axial forces [N11 (=0.08 and N=300kN) and
N19 (=0.09 and N=600kN)], greater strength degradation was observed
for higher levels of axial loading. This is explained by the mechanisms that
worsen the strength degradation, i.e., the concrete spalling and the
longitudinal reinforcement buckling.
a) column section 30x40m
2
b) column section 30x50m
2
c) column section 30x30m
2
Figure 53 Normalized strength degradation (2
nd
and 3
rd
cycle) for different load paths
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor X direction
N05 2nd cycl_X
N05 3rd cycl_X
N07 2nd cycl_X
N07 3rd cycl_X
N08 2nd cycl_X
N08 3rd cycl_X
N17 2nd cycl_X
N17 3rd cycl_X
2 4 6 8 10 12 14 16 18 20 22
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Y direction
N06 2nd cycl_Y
N06 3rd cycl_Y
N07 2nd cycl_Y
N07 3rd cycl_Y
N08 2nd cycl_Y
N08 3rd cycl_Y
N17 2nd cycl_Y
N17 3rd cycl_Y
1 2 3 4 5 6 7 8 9 10 11 12
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Xdirection
N09 2nd cycl_X
N09 3rd cycl_X
N11 2nd cycl_X
N11 3rd cycl_X
N12 2nd cycl_X
N12 3rd cycl_X
N18 2nd cic_X
N18 3rd cic_X
1 2 3 4 5 6 7 8 9 10 11 12
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Y direction
N10 2nd cycl_Y
N10 3rd cycl_Y
N11 2nd cycl_Y
N11 3rd cycl_Y
N12 2nd cycl_Y
N12 3rd cycl_Y
N18 2nd cic_Y
N18 3rd cic_Y
1 2 3 4 5 6
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor X direction
N13 2nd cycl_X
N13 3rd cycl_X
N14 2nd cycl_X
N14 3rd cycl_X
N15 2nd cycl_X
N15 3rd cycl_X
N16 2nd cycl_X
N16 3rd cycl_X
1 2 3 4 5 6
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Y direction
N13 2nd cycl_X
N13 3rd cycl_X
N14 2nd cycl_Y
N14 3rd cycl_Y
N15 2nd cycl_Y
N15 3rd cycl_Y
N16 2nd cycl_Y
N16 3rd cycl_Y
80 Chapter 4
a) column section 30x50m
2
b) column section 30x30m
2
Figure 54 Normalized strength degradation (2
nd
and 3
rd
) cycle for different levels of axial
load
4.5 Stiffness degradation
Stiffness degradation was evaluated by comparing the peaktopeak secant stiffness
values resulting from the first cycle of each imposed peak displacement (see Figure
55). Figure 56 and Figure 57 show the lateral peaktopeak stiffness degradation for
each column in each direction. From the results, the following can be concluded:
 Stiffness degradation evolves similarly under the uniaxial and the biaxial
demands in the two independent directions (Figure 56).
 The loading path history does not significantly influence the evolution of
the stiffness degradation (Figure 56).
 From the comparisons based on the experimental results, neither the axial
load ratio (ranging from 4 to 12%) nor the axial force level appear to
influence the stiffness degradation (Figure 57). However, other authors [139,
140], who studied columns with larger axial loads, found a more pronounced
strength degradation which started at lower amplitudes of lateral
displacement.
1 2 3 4 5 6 7 8 9 10 11 12
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor X direction
N11 2 cic_X (N=300kN v=0.08)
N11 3 cic_X (N=300kN v=0.08)
N19 2 cic_X (N=600kN v=0.09)
N19 3 cic_X (N=600kN v=0.09)
N20 2 cic_X (N=300kN v=0.045)
N20 3 cic_X (N=300kN v=0.045
1 2 3 4 5 6 7 8 9 10 11 12
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Xdirection
N11 2 cic_Y (N=300kN v=0.08)
N11 3 cic_Y (N=300kN v=0.08)
N19 2 cic_Y (N=600kN v=0.09)
N19 3 cic_Y (N=600kN v=0.09)
N20 2 cic_Y (N=300kN v=0.045)
N20 3 cic_Y (N=300kN v=0.045
1 2 3 4 5 6 7
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor X direction
N14 2 cic_X (N=210kN v=0.1)
N14 3 cic_X (N=210kN v=0.1)
N15 2 cic_X (N=210kN v=0.1)
N15 3 cic_X (N=210kN v=0.1)
N23 2 cic_X (N=650kN v=0.2)
N23 3 cic_X (N=650kN v=0.2)
N24 2 cic_X (N=650kN v=0.2)
N23 3 cic_X (N=650kN v=0.2)
1 2 3 4 5 6 7
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
X
Y
N
o
r
m
a
l
i
z
e
d
s
t
r
e
n
g
t
h
d
e
g
r
a
d
a
t
i
o
n
Displacement ductility factor Y direction
N14 2 cic_X (N=210kN v=0.1)
N14 3 cic_X (N=210kN v=0.1)
N15 2 cic_X (N=210kN v=0.1)
N15 3 cic_X (N=210kN v=0.1)
N23 2 cic_X (N=650kN v=0.2)
N23 3 cic_X (N=650kN v=0.2)
N24 2 cic_X (N=650kN v=0.2)
N23 3 cic_X (N=650kN v=0.2)
Analysis of Experimental Test Results 81
Consequently, regardless of the displacement history (i.e., uniaxial loading or any
of the biaxial loading paths), it appears that a simple relationship for stiffness
degradation can be used in simplified numerical models. Additionally, the relations
that have been proposed and calibrated for uniaxial models can be adopted for
biaxial models.
Figure 55 Stiffness degradation
a) 20x40cm
2
columns
b) 30x40cm
2
columns
i
s
K
1 + i
s
K
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB01N01
PB12N03
PB12N04
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB02N02
PB12N03
PB12N04
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB02N06
PB12N07
PB12N08
PB12N17
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB02N06
PB12N07
PB12N08
PB12N17
82 Chapter 4
c) 30x50cm
2
columns
d) 30x30cm
2
columns
Figure 56 Stiffness degradation for different load paths
a) 30x50cm
2
columns
b) 30x30cm
2
columns
Figure 57 Stiffness degradation for different levels of axial load
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB12N09
PB12N11
PB12N12
PB12N18
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB02N10
PB12N11
PB12N12
PB12N18
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB01N13
PB12N14
PB12N15
PB12N16
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB01N13
PB12N14
PB12N15
PB12N16
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB12N11 (N= 300; u=0.080)
PB12N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB12N11 (N= 300; u=0.080)
PB12N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift X (%)
PB01N14 (N= 210; u=0.1)
PB12N15 (N= 210; u=0.1)
PB12N23 (N= 650; u=0.2)
PB12N24 (N= 650; u=0.2)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.0
0.2
0.4
0.6
0.8
1.0
X
Y
R
e
l
a
t
i
v
e
s
t
i
f
f
n
e
s
s
Drift Y (%)
PB01N14 (N= 210; u=0.1)
PB12N15 (N= 210; u=0.1)
PB12N23 (N= 650; u=0.2)
PB12N24 (N= 650; u=0.2)
Analysis of Experimental Test Results 83
4.6 Dissipated energy
The energy dissipation is a fundamental structural property of RC elements when
subjected to seismic demands. For RC structures designed to accommodate damage
without collapse due to a seismic event, the input energy can be dissipated through
RC elements hysteretic response, without a significant reduction in strength [141].
Nonlinear static methods, for assessment or design, use energy dissipation capacity
related parameters to evaluate the inelastic earthquake response of structures and
to describe the strength and stiffness degradation of RC elements subjected to
cyclic loading [142].
4.6.1 Cumulative dissipated energy
Bousias et al. [48] stated that the strong coupling between the two transverse
directions of columns with biaxial loading produces an apparent reduction of
strength and stiffness in each of the two transverse directions when considered
separately, but also an increase in the hysteretic energy dissipation. This increase is
due to the larger width of the hysteresis loops in the transverse direction in the
presence of a nonzero force or deflection in the orthogonal direction.
Qiu et al. [22] claim that the cumulative hysteretic dissipation energy of a specimen
under biaxial loading is apparently larger than that under unidirectional loading
and is closely related to the loading position and path length.
The cumulative hysteretic dissipation energy was evaluated for all the tests,
considering the area of each loading cycle in the X and Y direction, and then the
total energy was calculated as the sum of these two parts, according to
Equations (1) to (3).
EJ
X
= _F
X
J
X
(1)
EJ
= _F
(2)
EJ
tot
= _F
X
J
X
+_F
(3)
The results in terms of evolution of cumulative dissipated energy are presented in
Figure 58 for the uniaxial and biaxial tests. In Figure 58, for each displacement
amplitude level, the plotted value of dissipated energy corresponds to the end of
the third cycle. For the quadrangular load path, the maximum displacement for
each cycle occurs in the path corner. These plots include also an additional series
(dashed line) representing the sum of the dissipated energy in each of the uniaxial
84 Chapter 4
tests for the corresponding column crosssection. From the analysis of the results
the following conclusions can be drawn:
 Comparing the two uniaxial test results, as expected, lower energy
dissipation was observed for the tested columns in its weakest direction,
associated with the lower column strength in this direction (15 to 20%
lower in the column with 30x40cm
2
section, and 60 to 80% in the column
with 30x50cm
2
section).
 The biaxial load paths induce larger amounts of dissipated energy than the
correspondent uniaxial paths. However, the sum of the dissipated energy in
the two unidirectional tests, the X and Y directions, leads to dissipation
energy evolution very close to that derived from the tests with rhombus
and circular load paths.
 The results in terms of dissipated energy evolution for the cruciform,
rhombus and circular load paths, is similar for all the columns tested. The
circular load path induces energy dissipation approximately 20% larger than
the rhombus load path. Rhombus load paths dissipate 30% more energy
than cruciform load paths.
 The quadrangular load path dissipates less energy than the other biaxial
load paths. It should be recalled that the maximum drift demand on the
quadrangular load path is reached in the path corner, corresponding to 2
times the maximum drift reached along the X and Y axes. Accordingly, the
quadrangular load path dissipates 30 to 45% less energy when compared to
the rhombus load path. However, the quadrangular load path would
dissipate 40 to 60% more energy than the rhombus load path.
 Comparing the dissipated energy of the biaxial load paths with the sum of
the dissipated energy in the two unidirectional tests, the rhombus load path
tends to dissipate more than 10 to 20%, and the circular load path
dissipates more than 20 to 40%. The lower bound of these differences is
found for the column with the square crosssection. This allows concluding
that, in the assessment and design of RC structures, neglecting the bending
interaction between each direction in the numerical models can introduce
an error of about 10 to 40% in terms of energy dissipation.
Analysis of Experimental Test Results 85
Figure 58 Comparison of cumulative dissipated energy for columns with different load
paths (uniaxial and biaxial loads)
4.6.2 Individual cycle energy
For a given column, both the energy dissipated in each individual loading cycle and
the accumulated energy dissipation along the whole test history were calculated.
For correlation purposes, particular cycles were also identified for which relevant
damage states occurred during the tests, namely the reinforcement bar buckling,
the conventional column failure and bar fracture. Examples of the obtained
graphics are presented in Figure 59 and Figure 60. From the analysis of the results
obtained for the 24 tested columns, the following can be concluded:
 In the 1
st
cycle of each peak displacement, larger energy dissipation is
observed relative to the subsequent cycles with the same peak displacement.
In the uniaxial tests, the reduction of the dissipated energy in the 2
nd
and
3
rd
cycles is about 10% of the energy dissipated in the 1
st
cycle. This
reduction is more pronounced (reaching 25%) for the biaxial load path. The
damage induced during the 1
st
cycle reduces the stiffness and strength, thus
decreasing the energy dissipation capacity of the column in the 2
nd
and 3
rd
cycles (see examples in Figure 59 and Figure 60).
 A significant drop in the energy dissipation is observed after reaching the
conventional failure of the column. This effect is associated with the
longitudinal bars buckling, which induces a high level of column strength
degradation.
0 1 2 3 4
0
25
50
75
100
125
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Maximum drift (%)
PB01N01
PB02N02
PB12N03
PB12N04
0 1 2 3 4 5
0
50
100
150
200
250
300
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Maximum drift (%)
PB01N05
PB02N06
PB12N07
PB12N08
PB12N17
PB01N05+PB02N06
0 1 2 3 4 5
0
50
100
150
200
250
300
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Maximum drift (%)
PB01N09
PB02N10
PB12N11
PB12N12
PB12N18
PB01N09+PB02N10
0 1 2 3 4 5
0
50
100
150
200
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Maximum drift (%)
PB01N13
PB02N14
PB12N15
PB12N16
PB01N13x2
86
F
Figu
igure 59 In
ure 60 Indi
ndividual cycl
ividual cycle
le energy and
columns (N0
energy and c
(N1
d cumulative
09 to N12 and
cumulative d
13 to N16)
dissipated en
d N18)
issipated ene
nergy for rect
rgy for squar
Chapter 4
tangular
re columns
Analysis of Experimental Test Results 87
4.6.3 Total dissipated energy until conventional collapse
According to Ohno and Nishioka [69] the total dissipated energy of a RC column
is independent of the loading path. This finding is in agreement with that of
Tsuno and Park [66]. However, in both studies the columns tested were all square
columns (40x40cm
2
and 55x55cm
2
respectively) and with axial load stresses
between 0.98 and 1.96MPa.
Figure 61 and Figure 62 compare the total dissipated energy obtained from the test
results. This total dissipated energy corresponds to the energy dissipated from the
start of the test until conventional rupture is reached, to a strength decay of 20%
relative to the maximum strength [143]. From the analysis of the results, the
following observations can be drawn:
 For square columns (N13 to N16) tested with a axial load stress
of 2.33MPa, the results obtained are in agreement with those reported by
Ohno and Nishioka [69] and by Tsuno and Park [66], i.e. the dissipated
energy until conventional rupture is approximately the same (see Figure
61d) with differences lower than 10%. However, for square columns (N23
and N24) with a higher level of axial load stress of 7.33MPa this conclusion
is not valid (see Figure 61b). The increase in axial load stress influences the
total energy dissipated.
 For rectangular columns, the finding of Ohno and Nishioka [69] is not valid
(see Figure 61a to Figure 61c), since the differences in strength and stiffness
of the two orthogonal directions induce differences that cannot be
dissociated from the biaxial coupling effect.
 Uniaxial tests in rectangular columns, for the same loading (comparing N05
with N06 and N09 with N10), show that the dissipated energy until
conventional rupture is dependent of the loading direction. The total
dissipated energy in the column weaker direction at the rupture point
is 90% (N06) and 20% (N10) huge than the corresponding results for the
tests in the strong direction (N05 and N09).
 The axial load ratio shows to directly influence the total energy dissipation
while the level of axial load force does not. In fact the response of similar
rectangular columns with different concrete grades tested for similar axial
load ratios [N11 (=0.08 and N=300kN) and N19 (=0.09 and N=600kN)]
result in dissipation energy that is approximately the same as the total
energy. On the other hand, similar rectangular columns with different
concrete grades tested for equal axial loading [N11 (=0.08) and N20
(=0.045) both with N=300kN] demonstrated that the total energy
88


Figur
dissipate
Figure 6
For squa
axial loa
(=0.2)
of axial
Figure 6
The non
columns
63), inc
conventi
re 61 Evalu
ed is higher
1).
are columns
ads [N14 a
with N=65
load ratio
1 and Figur
nrepetition
N21 and N
rease the
onal ruptur
a)
c)
uation of tota
l
r for the co
s tested wi
and N15 (
0kN] simila
present lo
re 62).
of the cyc
22 when com
total dissip
e.
al energy diss
loads (with d
olumn with
ith different
=0.1) with
ar findings w
ower values
cles, for the
mpared wit
pated energ
sipated of col
different load
h the lower
t axial load
h N=210kN
were establis
of total d
e same loa
h columns N
gy (10 to
lumns tested
paths)
axial load
d ratios an
N and N23
shed, i.e. hi
dissipated e
d path con
N7 and N8
20%) unti
b)
d)
for uniaxial
Chapter 4
ratio (see
nd different
3 and N24
igher levels
energy (see
nsidered in
(see Figure
il reaching
and biaxial
Analysis of E
Figure 6
Figure 6
4.6.4 N
du
As stated
displaceme
both facto
energy dis
the first yi
Figure 64
the corresp
uniaxial an
The bestf
shown in F
E
cum
E
= u
Experimental T
a)
2 Evaluatio
63 Evaluati
ormalize
uctility
by Elmens
ent ductility
ors to the
ssipation evo
ield point (E
shows the e
ponding dis
nd biaxial te
fit power co
Figure 64 an
u.64p
2.1
Test Results
on of total en
ion of total en
ed dissip
shawi and B
y and dissip
element va
olution was
E
y
).
evolution of
splacement
ests are repr
orrelation c
nd is given b
nergy dissipat
nergy dissipa
repetitio
pated en
Brown [141
pated energ
ariables. Fo
s normalised
f the norma
ductility fo
resented wit
curve for all
by Equation
ted for colum
ated for colum
on
nergy vs
], the relati
gy is comple
or each tes
d with the t
alized dissipa
r the tested
th different
l tests resu
n (4).
b)
mns with diffe
mns with and
displace
ion between
ex due to t
ted column
total dissipa
ated energy
d columns. R
mark filling
lts (uniaxia
erent axial lo
d without cyc
ement
n a RC ele
the sensitiv
n, the calcu
ated energy
y as a functi
Results from
g.
al and biaxi
(4)
89
oads
cle
ment'
ity of
ulated
until
ion of
m the
ial) is
90 Chapter 4
This expression is very similar to that obtained by bestfit correlation to the test
results from the uniaxial and biaxial loading separately.
As given by the proposed equation, for a displacement ductility of 4 (corresponding
to the minimum required ductility to withstand a severe earthquake), the
corresponding normalised dissipated energy estimated is 12.
Elmenshawi and Brown [141] and Nmai and Darwin [144] have investigated this
relationship for beams (with zero axial force), proposing similar equations. From
this expression a value of normalised dissipated energy for the same displacement
ductility is 3 times higher, around 35. This difference can be associated with the
axial loading levels.
As stated by Darwin and Nmai [145], the proposed equations need to be verified
with other experimental results. A validated expression can be very useful to
estimate the dissipated energy in the seismic design of RC elements in accordance
with international codes such as ACI 31808 [146].
Figure 64 Normalised dissipated energy vs displacement ductility
0 2 4 6 8 10 12
0
25
50
75
100
125
150
E
cum
/E
y
(=4) = 12
Uniaxial tests
Biaxial tests
Best fit curve
N
o
r
m
a
l
i
z
e
d
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
E
c
u
m
/
E
y
Displacement ductility
E
cum
/E
y
= 0.64
2.1
R
2
= 0.9617
Analysis of Experimental Test Results 91
4.7 Equivalent viscous damping ratio
4.7.1 Evaluation of equivalent damping from
experimental results
The equivalent damping depends on the structural displacement ductility demand
and the location of the plastic hinges in the elements [147]. The equivalent
damping is a sum of the elastic and hysteretic damping given by Equation (5),
were the hysteric damping (
hyst
) depends on the hysteretic characteristics of the
element. Elastic damping (
el
) is typically considered to be 5% for RC structures,
related to the critical damping [134].
cq
=
cI
+
hst
(5)
The equivalent hysteretic damping coefficient (
hyst
) can be given by Equation (6),
where E
D
and A
loop
are the shaded area shown in Figure 65, corresponding to the
dissipated energy; where E
S0
is the strain energy, and F
max
and D
max
are the
maximum force and displacement achieved in the loop.
hst
=
E
4nE
S0
=
A
Ioop
2nF
mux
mux
(6)
This procedure is accurate if the hysteretic response is perfectly symmetric for a
closed loop. In the case of seismic loads or even in the performed displacement
controlled tests, some asymmetries can be observed and the loops are not closed.
The equivalent damping can be evaluated for each halfcycle of the curves force
displacement according to the procedure proposed by Varum [24], and described
next:
 First, each halfcycle is identified, delimited by each pair of zeroforce
points;
 For each forcedisplacement halfcycle the maximum generalized force (F
max
)
and the maximum generalized displacement (D
max
) are evaluated, which
allows the calculation of the strain energy (E
S0
);
 For each halfcycle, the dissipated energy (E
D
) is computed by performing
the integral of the forcedisplacement curve;
 Finally, the equivalent hysteretic damping ratio (
eq
) is computed according
to Equation (5), adopting the Equation (7), for each halfcycle.
92 Chapter 4
hst
=
1
n
A
huI]Ioop
F
mux
mux
(7)
Figure 65 Damping for a hysteretic halfcycle
For each tested column, with a uniaxial or biaxial load path, the equivalent
damping was calculated, according to the presented methodology, for each
independent direction (X and Y) from the sheardrift curves. Subsequently a best
fit logarithmic curve was adjusted for each tested column in terms of equivalent
damping as a function of maximum ductility demand (see an example in Figure 66
for column N13).
Figure 66 Equivalent damping vs. maximum ductility demand and bestfit logarithmic
curves from column N13
Figure 67 to Figure 69 allow comparing the bestfit logarithmic curves obtained for
each tested column and each direction (X and Y) while the equations and the
correlation factors (R
2
) are summarized in Table 16. In particular, Figure 67 shows
the results for different load paths (uniaxial and biaxial) and Figure 68 reviews the
F
D
0 S
E
max
F
max
D
D
E
loop half
A
1 2 3 4 5 6 7
0
5
10
15
20
25
30
35
40
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
y = 6.1516ln(x) + 11.827
R = 0.956
PB01N13
X
Y
Analysis of Experimental Test Results 93
influence of the axial load and axial load ratio on the damping. Finally, Figure 69
focuses on the influence of the cycle repetition on damping. From the results
analysis of the in terms of the equivalent damping function of ductility it is
possible to verify that:
 The logarithmic bestfit curve correlates well with test results for the
uniaxial tests, but lower correlation factors were found from biaxial test
results (see Table 16).
 A significant influence of the load path on the equivalent damping was
found (see Figure 67). Generally, the biaxial load path induces higher
equivalent damping values when compared with the uniaxial tests.
 The quadrangular load path presents a higher equivalent damping than
that obtained from the other biaxial load paths.
 From Figure 68 it can be concluded that higher values of axial load tend to
induce larger equivalent damping in the strong direction (X), but no
influence was found for the column weak direction (Y).
 The equivalent damping results are compared in Figure 69, for biaxial
rhombus and quadrangular tests with and without repetition of cycles, for
the same axial load ratio, showing that the obtained results are very
similar.
94 Chapter 4
Figure 67 Bestfit equivalent damping vs. maximum ductility demand for different load
paths
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
X Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N01
PB12N03
PB12N04
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
Y Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB02N02
PB12N03
PB12N04
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
X Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N05
PB12N07
PB12N08
PB12N17
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
Y Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N06
PB01N07
PB01N08
PB01N17
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
X Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N09
PB12N11
PB12N12
PB12N18
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
Y Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB02N10
PB12N11
PB12N12
PB12N18
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X Direction
X
Y
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N13
PB12N14
PB12N15
PB12N16
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
Y Direction
X
Y
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N13
PB12N14
PB12N15
PB12N16
Analysis of Experimental Test Results 95
Figure 68 Bestfit equivalent damping vs. maximum ductility demand for different levels
of axial force
Figure 69 Bestfit equivalent damping vs. maximum ductility demand for the same load
path with and without cyclic repetition
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
X Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N11 (N= 300; u=0.080)
PB02N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
Y Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB01N11 (N= 300; u=0.080)
PB02N19 (N= 600; u=0.090)
PB12N20 (N= 300; u=0.045)
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
X Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB02N07 (N= 300; u=0.12)
PB02N08 (N= 300; u=0.12)
PB12N21 (N= 620; u=0.12)
PB12N22 (N= 620; u=0.12)
0 2 4 6 8 10
0
5
10
15
20
25
30
35
40
45
50
X
Y
Y Direction
E
q
.
D
a
m
p
i
n
g
(
%
)
Displacement ductility factor
PB02N07 (N= 300; u=0.12)
PB02N08 (N= 300; u=0.12)
PB12N21 (N= 620; u=0.12)
PB12N22 (N= 620; u=0.12)
96 Chapter 4
Table 16 Best fit logarithmic curve for global damping for tested each column
Column Specimen dir Best Fit Curve R
2
2
0
x
4
0
P01N01 X
cq
= 8.17ln(p) + 6.25 0.94
P02N02 Y
cq
= 3.08ln(p) + 13.09 0.54
P12N03
X
cq
= 6.41ln(p) + 8.65 0.92
Y
cq
= 3.14ln(p) + 12.64 0.39
P12N04
X
cq
= 5.79ln(p) + 8.92 0.89
Y
cq
= 1.28ln(p) + 9.19 0.14
3
0
x
4
0
P01N05 X
cq
= 6.96ln(p) + 6.01 0.84
P02N06 Y
cq
= 4.55ln(p) + 9.32 0.57
P12N07
X
cq
= 7.451ln(p) + 10.39 0.86
Y
cq
= 3.44ln(p) + 9.49 0.56
P12N08
X
cq
= 6.16ln(p) + 16.83 0.58
Y
cq
= 3.73ln(p) + 18.68 0.42
P12N17
X
cq
= 12.17ln(p) + 12.14 0.64
Y
cq
= 9.49ln(p) + 7.61 0.61
3
0
x
5
0
P01N09 X
cq
= 6.06ln(p) + 6.88 0.85
P02N10 Y
cq
= 7.03ln(p) + 7.23 0.95
P12N11
X
cq
= 8.57ln(p) + 5.85 0.89
Y
cq
= 4.86ln(p) + 7.41 0.81
P12N12
X
cq
= 7.51ln(p) + 14.49 0.81
Y
cq
= 10.43ln(x) + 15.13 0.82
P12N18
X
cq
= 5.83ln(p) + 14.23 0.46
Y
cq
= 11.50ln(p) + 0.679 0.75
3
0
x
3
0
P01N13 X
cq
= 6.15ln(p) + 11.83 0.96
P12N14
X
cq
= 9.39ln(p) + 7.62 0.89
Y
cq
= 9.98ln(p) + 3.57 0.79
P12N15
X
cq
= 10.04ln(p) + 19.25 0.69
Y
cq
= 8.18ln(p) + 15.48 0.87
P12N16
X
cq
= 3.465ln(p) + 21.86 0.31
Y
cq
= 9.42ln(p) + 8.37 0.75
3
0
x
5
0
P12N19
X
cq
= 10.41ln(p) + 6.42 0.87
Y
cq
= 1.95ln(p) + 12.05 0.19
P12N20
X
cq
= 6.51ln() + 7.59 0.80
Y
cq
= 4.53ln() + 7.99 0.69
3
0
x
4
0
P12N21
X
cq
= 5.64ln() + 10.94 0.74
Y
cq
= 4.59ln(p) + 8.81 0.52
P12N22
X
cq
= 7.31ln(p) + 14.78 0.59
Y
cq
= 6.39ln(p) + 15.61 0.54
Analysis of Experimental Test Results 97
4.7.2 Empirical proposals for equivalent damping in RC
elements under uniaxial loadings
Different proposals for the equivalent damping of RC elements and structures can
be found in the literature, including the works of Priestley et al. [148, 149],
Rosenblueth and Hererra [150], Gulkan and Sozen [151], Kowalsky [152], among
others. Bandon [153] developed an extensive review and study of the existing
proposals for all type of elements. The more frequently used equations for
predicting the equivalent damping () of RC columns as a function of ductility (p)
are the following:
Rosenblueth and Hererra [150], based on the response of a bilinear elastoplastic
system, where r is the post yield stiffness coefficient,
cq
=
0
+
2
n
_
(1 r)(p  1)
p  rp + rp
2
_ (8)
Gulkan and Sozen [151], based on the Takeda model,
cq
=
0
+u.2 _1 
1
p
] (9)
Kowalsky [152], based on the Takeda model with a =0.5 and =0, where r is the
post yield stiffness coefficient,
cq
=
0
+
1
n
_1 
1 r
p
 rp]
(10)
Stojadinovic and Thewalt [154], developed the following expression based on
regression analysis for experimental results of columns and beams tested under
quasistatic cyclic loading conditions:
cq
= _
4.7 p < 1.u
u.4p
2
+ 7.1p 2 otbcrwisc
(11)
Lu et al. [155], based their formula on shaking table test results for RC frames:
cq
= 1uu  6.S(p  S)
2
for p < S.u
(12)
Priestley [148], for concrete columns and walls:
cq
= S +
9S
n
_1 
1
p
]
(13)
Dwairi and Kowalsky [156], where C is dependent on the hysteresis rule
98 Chapter 4
cq
= C _
p 1
pn
]
(14)
Priestley et al. [157], for concrete framed buildings
cq
= u.uS +u.S6S _
p  1
pn
]
(15)
The results given by the presented equations were compared with the experimental
results obtained from the uniaxial tests. In order to evaluate the accuracy of the
presented empirical expressions, the correlation factor between the predicted values
and the experimental results has been calculated. The correlation is analysed
individually for each tested columns (see Figure 70) and also for all test results (see
Figure 71). From an analysis of the results, using the uniaxial test data the
proposals of Kowalsky [152], Priestley [148] and Priestley et al. [157] show better
correlation with the experimental results (R
2
>0.7), as evidence in Figure 72.
Legend: RH Rosenblueth and Hererra [150]; GS  Gulkan and Sozen; K  Kowalsky[152];
ST  Stojadinovic and Thewalt [154]: LHCM  Lu et al. [155]; P  Priestley [148]; DK 
Dwairi and Kowalsky [156]; PCK  Priestley et al. [157])
Figure 70 R
2
results for the equation proposals for equivalent damping and each uniaxial
test result
From the comparative analysis of the results in Figure 70 and Figure 71 it was
observed that, apart from the expression proposed by Gulkan and Sozen [151] and
Lu et al. [155], the others give correlation factors for the overall tests similar to the
average of the correlation factor obtained for each individual test. This is justified
by the larger dispersion between the correlation factors obtained from both
expressions for each individual test.
RH GS K ST LHCM P DK PCK
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
R
2
Equations for the equivalent damping
PB01N01
PB02N02
PB01N05
PB02N06
PB01N09
PB02N10
PB01N13
Analysis of E
Legend: R
ST  Sto
Figure 71
Figure 72
4.7.3 Eq
The equiv
effective d
respective
cq
=
x
where
cq
equivalent
E
y
are the
Experimental T
RH Rosenblu
ojadinovic and
Dwai
1 R
2
results
2 Equivalen
quivalen
valent damp
damping es
potential en
E
x
+
E
E
x
+E
is the equ
t damping e
potential en
R
2
Test Results
ueth and Her
d Thewalt [1
ri and Kowal
for the equa
nt damping e
nt biaxia
ping was com
timated for
nergy, given
E
uivalent dam
estimated fo
nergy in eac
RH GS
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
E
rerra [150]; G
154]: LHCM 
lsky [156]; PC
ation proposa
test resu
estimated wit
uniaxial t
al dampi
mputed for
r each inde
n by Equatio
mping of th
or each indiv
ch direction
K ST
Equations for the eq
GS  Gulkan a
 Lu et al. [15
CK  Priestle
als for equival
ult
th empirical e
tests
ng
each biaxia
ependent d
on (16):
he biaxial r
vidual direc
.
LHCM P
quivalent damping
and Sozen; K
55]; P  Pries
ey et al. [157]
lent damping
expressions a
al test, as a
irection we
response;
x
tion (X and
DK PCK
g
K  Kowalsky
stley [148]; D
]
g and all unia
and results fo
a function o
eighted with
x
and
y
ar
d Y); and E
99
[152];
K 
axial
r all
of the
h the
(16)
re the
E
x
and
100 Chapter 4
Figure 73 shows the obtained global equivalent damping, according to the given
methodologies, for each load path type. The empirical expressions with better
correlation to the uniaxial test results are also highlighted. From the analysis of
Figure 73, the following can be concluded:
 The estimated equivalent global damping for biaxial tests is clearly
dependent on the load path. For example, for a ductility factor of 2 the
rhombus load path has an equivalent damping of around of 10%, while
quadrangular and circular paths have an equivalent damping of 20%.
 Comparing the results obtained for uniaxial and biaxial load paths, the
cruciform and rhombus paths present similar equivalent global damping.
However, the quadrangular and circular paths evidence higher levels of
damping when compared to that obtained from the uniaxial tests. For
example, for a ductility factor of 6 the uniaxial load paths induce an
equivalent damping of around 20%, while for the quadrangular a value of
about 30% is observed.
From the previous comments, it is concluded that the typically used equations to
estimate equivalent damping for uniaxial stress cannot be used for equivalent
damping in biaxial loading conditions. However, for displacement biaxial paths
close to the cruciform and rhombus the empirical equations exhibit acceptable
results.
From the analysis of test results of 46 columns, Fardis and Panagiotakos [158] have
confirmed that biaxial loading achieves higher values of hysteretic damping when
compared to uniaxial loading. Bousias et al. [48] stated that the larger damping
observed for biaxial loading is due to the coupling response of the columns between
the two transverse directions.
Despite the expressive dispersion of the viscous damping calculated from biaxial
tests, different equations were adjusted, by fitting the experimental data. The two
bestfitting equations are included in Figure 74 (Equation (17) and (18)), where
both equations have poor a correlation factor relative to the experimental results
(R
2
) of 0.31.
The obtained low correlation factor is justified by the dispersion of the equivalent
damping determined from biaxial tests, is particularly dependent on the load path.
However, the proposed equation can be considered as a first estimation of the
equivalent damping for RC columns under biaxial loading.
cq
= SS.6 
22.2S
p
0.37
(17)
cq
= 12.S6 + S.18 ln p
(18)
Analysis of E
4.8 Ul
The recen
(i.e., the
seismic cap
Specifically
buildings
of existing
Experimental T
Fi
Figure 74
timate
nt European
Eurocodes
pacity of RC
y, Part 3
(EC83) [15
g structure
Test Results
igure 73 Eq
Best fitted
displac
n standards
and the n
C buildings
of Eurocod
59] describes
s [160]. Th
quivalent dam
d proposals fo
cement
for the ea
national ann
with nonlin
de 8 regard
s provisions
he use of
mping for bia
or biaxial equ
capacit
arthquakere
nexes) cons
near method
ding the st
for the seis
nonlinear
axial tests
uivalent dam
ty
esistant desi
sider the ev
dologies.
trengthening
smic perform
methodolog
mping
ign of struc
valuation o
g and repa
mance evalu
gies require
101
ctures
of the
air of
uation
s the
102 Chapter 4
estimation of the postelastic properties of RC elements, such as the yield strength,
yield deformation and peak strength, among other properties [161].
Concerning the safety assessment at the member level, EC83 provides expressions
for both ductile and brittle mechanisms for the member capacity, defined in terms
of the admissible member chord rotation to be used for a specific limit state, thus
allowing for the available ductility estimation. Bora et al. [162] performed a series
of uniaxial tests on RC columns and concluded that the deformation limits
proposed by EC8 are very conservative in their estimation of rotation thresholds
for the life safety (i.e., resulting in significant damage) and collapse prevention
(i.e., resulting in near collapse) limit states when such thresholds are compared
with test results. Popa et al. [163] reached similar conclusions based on a
comparison of RC column test results and the EC8 provisions for the estimation of
ultimate displacement.
For the limit state corresponding to near collapse, EC83 provides two alternative
expressions for the evaluation of the ultimate element capacity of RC elements. In
particular, the ultimate rotation capacity under cyclic loading can be obtained by
Equation (19) [EC83 Equation (A.1)]:
0
um
=
1
y
cI
u.u16 (u.S)
v
_
mox (u.u1;
i
)
mox (u.u1; )
_
0.225
_
I
v
b
]
0.35
2S
_up
sx
]
jw
]c
]
(1.2S
100p
d
)
(19)
where el is 1.5 for primary members and 1.0 for secondary members; is the
normalised axial force; and are the mechanical reinforcement ratios for tension
and compression, respectively; fc is the concrete compressive strength; fyw is the
estimated value of the stirrup yield strength; sx = Asx/bwsh is the transverse steel
ratio (Asx is parallel to the direction of loading, sh is the stirrup spacing and bw is
the crosssection width); d is the diagonal reinforcement steel ratio in each
diagonal direction; LV is the shear span (considered constant and equal to half the
member length); h is the crosssection depth and is the confinement effectiveness
factor.
The ultimate rotation may also be calculated by an alternative expression based on
the evaluation of the ultimate plastic section curvature, assumed to be constant
over the plastic hinge length and obtained by Equation (20)
[EC83 Equation (A.4)]:
0
um
=
1
y
cI
_0
+ (
u

) I
pI
_1 u.S
I
pI
I
v
]_
(20)
Analysis of Experimental Test Results 103
where 
y
and 
u
are the yield and ultimate curvatures at the end section,
respectively, L
pl
is the plastic hinge length and
y
is the yield chord rotation as
defined by Equation (A.10) of EC83 .
Verderame et al. [161] concluded that these equations may not provide accurate
estimations of the ultimate rotation because they do not account for certain
parameters that might affect the rotational capacity, such as the load path and the
energy dissipation in hysteretic cycles. These limitations have been also
experimentally investigated by Pujol et al. [164].
Considering the geometric and mechanical characteristics of the RC tested
elements, the ultimate rotation was calculated for each crosssection type. The
values obtained from the two EC83 equations were compared (for each column
direction) with the ultimate drift estimated from the tests (corresponding to a
conventional 20% reduction in peak force). For comparing the ultimate drift
observed in the tests with the values given by the EC83 equations, it was assumed
that the ultimate drift corresponds to the value obtained for the columns strong
direction. The results obtained for the ultimate drift are presented in Figure 75.
From the analysis of the uniaxial test results, it can be concluded that the values
obtained with Equation (19) present good agreement with the experimental
rotation values for the strong direction. In the weak direction, this equation
underestimates the ultimate drift, while Equation (20) underestimates the results
for all of the tested columns.
As already noted, the biaxial loading test leads to lower ultimate drift of the
columns than does the corresponding uniaxial loading test. Equation (19)
provides ultimate capacity values larger than those experimentally observed,
whereas Equation (20) tends to underestimate the experimentally observed values
in most of the analysed cases. As stated by Rozman and Fajfar [165], there is still a
large uncertainty in the determination of the deformation capacity, depending on
the approach adopted. Additionally, Verderame et al. [161] stated that the large
coefficient of variation associated with these equations is due to the expected
variability in the experimental results, and also to the difficulty in completely
modelling the interactions between the complex phenomena influencing the post
elastic deformation behaviour of the RC elements with a simple formula.
In fact, as recognised by other authors [161, 162, 165], EC8 can lead to ultimate
deformation values lower than the experimental results for uniaxial demands. As
shown in Figure 75, lower values of the ultimate deformation demand were often
obtained from the EC8 equations, when compared with the experimental results.
This may be primarily due to the more pronounced strength degradation observed
under biaxial loading, with results in a lower ultimate ductility. Therefore, the use
of EC8 equations to estimate the ultimate deformation of columns subjected to
biaxial demands should be further investigated.
104 Chapter 4
Table 17 Ultimate drift ratios (experimental/analytical values according to EC83)
Specimen drift
u,exp
1
drift
u,EC8a
2,4
drift
u,exp
/
drift
u,EC8a
drift
u,EC8b
3,4
drift
u,exp
/
drift
u,EC8b
U
n
i
a
x
i
a
l
t
e
s
t
s
PB01N01 3.34 3.85 0.87 2.05 1.63
PB02N02 N/A 4.89 6.04
PB01N05 3.01 2.98 1.01 1.52 1.98
PB02N06 4.66 3.29 1.42 2.10 2.22
PB01N09 3.00 2.94 1.02 1.43 2.10
PB02N10 5.02 3.50 1.43 3.14 1.60
PB01N13 3.97 3.43 1.16 2.05 1.97
B
i
a
x
i
a
l
t
e
s
t
s
PB12N03 3.01 3.85 0.78 2.05 1.47
PB12N04 3.32 3.85 0.86 2.05 1.62
PB12N07 2.32 2.98 0.78 1.52 1.53
PB12N08 2.38 2.98 0.80 1.52 1.57
PB12N11 2.32 2.94 0.79 1.43 1.62
PB12N12 2.37 2.94 0.81 1.43 1.66
PB12N14 2.97 3.43 0.87 2.05 1.45
PB12N15 3.25 3.43 0.95 2.05 1.59
PB12N16 3.33 3.43 0.97 2.05 1.62
PB12N17 1.72 3.33 0.52 1.47 1.17
PB12N18 2.05 3.21 0.64 1.46 1.40
PB12N19 1.41 3.47 0.41 1.75 0.81
PB12N20 1.72 3.28 0.52 1.38 1.25
PB12N21 4.35 3.34 1.30 1.52 2.86
PB12N22 2.96 3.45 0.86 1.52 1.95
PB12N23 1.41 3.34 0.42 1.46 0.97
PB12N24 1.72 3.34 0.51 1.46 1.18
1
Ultimate drift verified experimentally corresponding to a drop of 20% of peak
resistance
2
Ultimate drift obtained with the Expression A.1 of the EC83
3
Ultimate drift obtained with the Expression A.3 of the EC83
4
Ultimate drift considered for the strong direction of the columns
Figure 75 Ultimate drift capacity: Experimental vs Analytical (EC83 expressions)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Uniaxial strong direction
Uniaxial weak direction
Biaxial
D
r
i
f
t
c
a
p
a
c
i
t
y
(
%
)

E
x
p
e
r
i
m
e
n
t
a
l
Drift capacity (%)  according to EC83 Eq. A.1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Uniaxial strong direction
Uniaxial weak direction
Biaxial
D
r
i
f
t
c
a
p
a
c
i
t
y
(
%
)

E
x
p
e
r
i
m
e
n
t
a
l
Drift capacity (%)  according to EC83 Eq. A.4
Analysis of Experimental Test Results 105
4.9 Damage quantification
The structural damage induced by earthquakes can be quantified in order to assess
the structure relative to a predefined set of limit states [38].
Damage indices are generally divided into two categories: local and global. Local
indices are calibrated for a specific structural member. By contrasts, global indices,
are used to predict the failure of a complete structure, and are usually computed as
a weighted combination of local damage indices of individual members [166].
A good damage assessment method should have general applicability, i.e., it should
be valid for a variety of structural systems, based on a simple formulation, be easy
to use, and generate easily interpretable results [24, 167].
Damage indices can be classified as cumulative or noncumulative. Noncumulative
indices relate the state of damage to peak response quantities and do not account
for cyclic loading effects. Cumulative indices include part or all of the loading
history to predict the capacity reduction due to cyclic repetitive loading [166].
Different authors have performed extensive reviews of the damage indices proposed
in the literature [33, 168170].
The formulation used in this study for the damage index estimation is based on the
methodology proposed by Park and Ang [143]. In the following section, this index
and the corresponding results for the columns tested uniaxially are presented. Also,
different combinations of the original Park and Ang methodology are adopted and
applied to columns tested biaxially.
4.9.1 Original Park and Ang damage index
The damage index proposed by Park and Ang [128, 143] is defined by a linear
combination of the normalised maximum deformation and the normalised
dissipated hysteretic energy resulting from cyclic loading. The damage index DI is
therefore expressed by:
I =
J
mux
J
u
+[
]JE
F
J
u
(21)
where d
max
is the maximum displacement of the structural member, d
u
is the
ultimate displacement, dE is the dissipated hysteretic energy, and F
y
is the
yielding strength of the structural member; is a degradation parameter which
represents the influence of cyclic response on column damage and can be estimated
with empirical expressions based on structural parameters. The damage index
106 Chapter 4
typically ranges from 0 to 1, since it is formally upper bounded by the unitary
value; however, when DI > 1, total damage (collapse) is assumed.
Several empirical expressions based on experimental results can be found in the
literature to estimate the strength degradation parameter (). A typical value
of 0.05 is often adopted [13]. Equation (22) was proposed by Kunnath et al. [171]
and is one of the most used expressions to estimate the parameter .
[ = u.9
100p
w
(u.S7 mox{v; u.uS] + u.S (
t
u.17)
2
)
(22)
where
w
is the volumetric confinement ratio (volume of closed stirrups divided by
the volume of confined concrete core), v is the normalised axial force (taken
positive for compression) and
t
is the mechanical ratio of tensile longitudinal
reinforcement.
Concerning the ultimate deformation capacity, some expressions have been
proposed based on experimental tests performed up to failure on beams and
columns. Park et al. [128] suggested Equation (23) to estimate the ultimate
displacement. According to the study performed by Fardis et al. [170], it was
concluded that this equation leads to the best agreement of the Park and Ang
index with several experimental results analysed.
R
u
(%) = 1.9S8 _
l
J
]
0.93
p
0.27
p
w
0.48
v
0.48
c
0.15
(23)
R
u
(%) =
o
u
l
; p = p
t
c
; v =
N
b J
c
(24), (25), (26)
R
u
(%) is the ultimate rotational capacity (in percentage); o
u
is the ultimate
horizontal displacement capacity;
I
d
is the shear span ratio; p is the normalised steel
ratio; p
t
is the volumetric ratio of longitudinal steel; p
w
is the confinement ratio (in
percentage; taken as 0.4% if p
w
< 0.4%); is the normalised axial force (taken as
0.05 if < 0.05); N is the axial load (in kN); b is the crosssection width; d is the
effective depth of the crosssection;
c
is the concrete strength (in kPa); and
is
the yielding strength of steel reinforcement (in kPa).
In order to establish a correspondence between the calculated damage index values
and the experimentally observed damage, Table 18 presents the damage index
ranges for each damage degree, based on postearthquake damage reports [24].
Analysis of Experimental Test Results 107
Table 18 Calculated damage index vs. observed damage
Damage inspection Calculated
local damage
index
Degree of damage Physical appearance
Collapse Total or partial building collapse >1.00
Severe
Extensive crushing of concrete. Disclosure of
buckled reinforcements
0.751.00
Moderate
Extensive large cracks. Spalling of concrete
in weaker elements
0.350.75
Minor
Minor cracks throughout building. Partial
crushing of concrete columns
0.100.35
Slight Sporadic occurrence of cracking 0.000.10
4.9.2 Application of Park and Ang damage index to the
uniaxial tests
The damage index was computed from the uniaxial test results. Equation (22) was
used to estimate , for each direction of RC columns, and the ultimate
displacement was computed using Equation (23). The corresponding results (for
both directions) are summarized in Table 19.
Table 19 Estimated parameters for RC column damage index calculations
Column
o
u
(mm)
PB01N01 0.038 92.94
PB01N05 0.047 80.22
PB02N06 0.033 142.44
PB01N09 0.038 73.46
PB02N010 0.038 120.56
PB01N13 0.043 108.84
The results are plotted in Figure 76 in terms of DI evolution during the cyclic tests.
In the DI curves it is also presented, the corresponding drift for each damage state
observed in the tests and reported in the previous section.
108
Acco
colum
evolu
result
colum
In Fi
the D
drift
drift
the la
F
ording to th
mns uniaxia
ution curves
ts are in g
mns.
igure 77, th
DI is plotte
values the
demands th
ast cycles).
Figure 77
R
e
l
a
t
i
v
e
c
o
n
t
r
i
b
u
t
i
o
n
o
f
t
h
e
m
a
x
i
m
u
m
Figure 76 D
he results ob
ally tested,
s and the d
good agreem
he relative c
ed for all co
DI basically
he DI is hig
Contributi
0.0 0.5
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
d
e
f
o
r
m
a
t
i
o
n
c
o
m
p
o
n
e
n
t
t
o
t
h
e
D
I
Damage index
btained in t
the follow
amage desc
ment with t
ontribution
olumns teste
y depends o
ghly influen
ion of the ma
1.0 1.5
x evolution fo
terms of da
wing can b
cription asso
the physical
of the max
ed uniaxiall
on the maxi
ced by the
aximum defor
2.0 2.5 3.0
or the uniaxi
amage index
be conclude
ociated with
l damage o
ximum defor
ly. As can b
imum defor
dissipated e
rmation comp
3.5 4.0
al tests
x evolution
d. Based o
h each DI v
bserved in
rmation com
be observed
mation, but
energy (up
ponent to the
4.5 5.0
Drift (%)
PB01N01
PB01N05
PB02N06
PB01N09
PB02N10
PB01N13
Chapter 4
for the six
on the DI
value, these
the tested
mponent to
d, for lower
t for larger
to 50% for
e DI
Analysis of Experimental Test Results 109
4.9.3 Proposals of damage index for RC columns under
biaxial loading conditions
As previously discussed, RC elements biaxially loaded go through the damage
states for earlier levels of drift demand. As the damage indexes for RC elements
available in the literature (such as that proposed by Park and Ang, for example)
were mainly developed and calibrated for elements under uniaxial loading
conditions, the adequacy of recent proposals of damage index formulations for
columns under biaxial loading, as well as new proposals by the authors of this
work, was checked as described next.
All these proposals are based on the philosophy of the original Park and Ang
damage index model, adding one component for the maximum deformation demand
and another for the dissipated energy. The proposals are validated against test
results of columns loaded biaxially and their accuracy in representing the damage
evolution of RC elements subjected to biaxial loads is discussed.
Equations (27) to (30) for the evaluation of the damage index of RC columns under
biaxial loading (DI
biaxial
) were previously studied by Qiu et al. [22]. In this study,
the damage indexes given by these equations as well as by three other Equations
((31) to (33)) are assessed and compared. In these expressions, DI
biaxial
refers to the
biaxial damage index, D
x
and D
y
are the damage indexes calculated for each
independent direction, E
x
and E
y
are the cumulative dissipated energy calculated
for each independent direction, and is a constant parameter experimentally
calibrated.
In Equation (33), the total damage index due to biaxial loading is not calculated as
the sum of the damage indexes for each direction but is based on the calculation of
the resultant displacements and total energy. Thus, for each step, the resultant
maximum displacement (d
max,res
) is calculated, and for this resultant direction the
ultimate displacement (d
u,res
) and the equivalent resultant yielding force (F
y,res
) are
estimated. For each step, the total energy is simply calculated as the sum of the
dissipated energy associated with each direction. The resultant ultimate
displacement and the resultant yielding force, evaluated in each step of the
analysis, are calculated for the loading direction (), based on a resultant curve
with an assumed shape. For this curve, four different shapes are tested, namely one
linear, two elliptic, and one parabolic, as shown in Figure 78. Of the four studied
curves, the strategy based on the parabolic one gives the best results, and so this
curve is adopted for the comparisons developed in the next section.
110 Chapter 4
I
buxuI
= I
x
+I
z min(I
x
, I
)
(27)
I
buxuI
= mox(I
x
, I
) +
E
x
or E
E
x
+ E
min(I
x
, I
)
(28)
I
buxuI
= mox(I
x
, I
) +
I
x
or I
I
x
+ I
min(I
x
, I
)
(29)
I
buxuI
=
`
1
1
1
1
J
mux,x
J
u,x
+[
E
x
+ E
F
,x
J
u,x
(J
mux,x
> J
mux,
)
J
mux,
J
u,
+[
E
x
+ E
F
,
J
u,
(J
mux,
> J
mux,x
)
(30)
I
buxuI
= I
x
+I
(31)
I
buxuI
= _I
x
2
+I
2
(32)
I
buxuI
=
J
mux,cs
J
u,cs
+ [
E
x
+ E
F
,cs
J
u,cs
(33)
Figure 78 Curves for evaluation of resultant displacement
4.9.4 Results
The accuracy of each expression presented in the previous section for the
estimation of damage in RC columns under biaxial loading conditions is evaluated
based on a scoring procedure. For each of the four damage states defined based on
visual inspection, the damage indexes calculated by each expression are assessed,
and a score is assigned. Thus, for each damage state, if the DI calculated with a
certain expression gives a value within the range presented in Table 20 it is
assigned a score of one; otherwise it is assigned zero. Therefore, for each column the
1
2
3
4
ult,x
ult,y
1
2
3
4
Linear
Ellipse
Ellipse
Parabolic
Analysis of Experimental Test Results 111
score may vary between 0 and 4. The range limits presented in Table 20 are
defined based on the observed damage states, as reported in previous sections.
Table 20 Damage index ranges for each damage state adopted in the scoring procedure
Damage state DI Range
Conventional rupture > 0.80
Reinforcing steel bar
buckling
0.550.80
Concrete spalling 0.300.70
Cracking 0.000.10
The damage index was computed for each biaxial test using Equations (27) to (33).
Equations (22) and (23) were adopted to estimate the parameter and the
ultimate displacement (for each direction of the RC column). The corresponding
results are summarized in Table 21.
Table 21 Estimated parameters for RC column damage index calculations
Colum
o
u,Stong
(mm) o
u,wcuk
(mm)
PB12N03
0.038 92.9 167.4
PB12N04
PB12N07
0.047 80.2 142.4
PB12N08
PB12N11
0.038 73.5 120.6
PB12N12
PB01N14
0.043 108.8 PB01N15
PB01N16
PB12N17 0.055 62.1 98.2
PB12N18 0.042 77.5 113.7
PB12N19 0.032 97.6 143.3
PB12N20 0.049 69.9 102.7
PB12N21
0.059 59.3 93.7
PB12N22
PB12N23
0.083 66.7
PB12N24
For the DI calculations according to Equation (27), the parameter proposed by
Qiu et al. [22] is equal to 0.5; however, for the experimental results analysed in this
work a value of =0.85 gives the highest score. For the proposal presented in
Equation (33), the best results are obtained for the parabolic curve, given by
Equation (34) and (35), dependent on the parameter P. Comparing with the
experimental results, a value of P=1.8 gives the best score.
112 Chapter 4
y = K +
1
4P
(x E)
2
(34)
with
K =
1
4
_4 o
uIt,

[6
ult,x
2
+46
ult,j
P
2
46
ult,x
2
P
_ , and
E =
1
4
_
o
uIt,x
2
+4 o
uIt,
P
2 o
uIt,x
_
(35)
The damage indexes estimated using the different expressions are summarized in
Table 22. The adequacy of each expression in representing the damage evolution is
finally assessed, summing the scores for each of the 15 columns studied. Equations
(27) and (33) yielded the best total scores. Figure 79 and 80 present the evolution
of the damage index calculated for each biaxial test of the RC columns with
Equations (27) and (33), respectively. The plots represent the damage states
observed in the tests (cracking, concrete spalling, bar buckling, and conventional
rupture). As can be observed, both expressions give damage indexes that agree well
with the physical damage evolution observed in the columns during the cyclic tests.
Table 22 Scores obtained for each column with the different biaxial DI equations
Equations
Column Eq. (27) Eq. (28) Eq. (29) Eq. (30) Eq. (31) Eq. (32) Eq. (33)
N7 4 4 4 2 3 4 2
N8 4 3 3 3 1 3 4
N11 3 4 4 2 3 4 2
N12 3 3 3 2 2 3 4
N14 3 1 1 2 1 2 3
N15 4 2 2 2 1 2 4
N16 4 4 4 4 2 4 4
N17 4 3 2 2 2 4 2
N18 3 3 3 3 4 3 2
N19 2 4 4 2 4 2 1
N20 3 2 2 2 1 2 4
N21 2 2 2 2 2 2 3
N22 3 2 2 3 1 2 2
N23 3 2 2 3 1 2 2
N24 2 1 1 1 1 1 3
Total 47 40 39 35 29 40 42
Analysis of Experimental Test Results 113
Figure 79 Damage index evolution calculated for each biaxial test with Equation (27)
Figure 80 Damage index evolution for the biaxial tests according with Equation (33)
4.10 Final remarks
This chapter presented the results obtained from an experimental campaign on
twentyfour flexuredominated RC columns, with different geometries and
reinforcements, that were subjected to uniaxial and biaxial horizontal displacement
paths with constant axial loading. From the results of this experimental
investigation, several conclusions can be drawn:
 For each tested column, the drift corresponding to the characteristic
damage was identified, in order to understand the influence of biaxial
loading on the damage evolution. Globally, it was observed that each
damage state for the biaxial loading conditions occurred for lower drift
demands than the corresponding damage state for uniaxial loading.
Concrete cracking, concrete spalling and bar buckling were observed in the
biaxial tests for a drift demand 5075% of the value for the corresponding
uniaxial tests. In the biaxial tests, conventional rupture was also observed
for lower drift demands, due to the pronounced strength degradation after
buckling of reinforcing bars.
0 1 2 3 4
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
A
B
A
B CD
A
B CD
A
B
C
D
Conventional rupture
Bar buckling
Spalling of concrete
Concrete cracking
PB12N7
PB12N8
PB12N11
PB12N12
PB12N14
PB12N15
PB12N16
PB12N17
PB12N18
PB12N19
PB12N20
PB12N21
PB12N22
PB12N23
PB12N24
Drift Level(%)
COLLAPSE
MODERATE
Extensive large cracks.
Spalling of concrete in
weaker elements
SEVERE
Extensive crushing of
concrete.
Disclosure of buckled
reinforcements
MINOR
Minor cracks.
Partial crushing of
concrete columns
SLIGHT
Sporadic cracking
0 1 2 3 4
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B CD
A
B CD
A
B
C
D
Conventional rupture
Bar buckling
Spalling of concrete
Concrete cracking
PB12N7
PB12N8
PB12N11
PB12N12
PB12N14
PB12N15
PB12N16
PB12N17
PB12N18
PB12N19
PB12N20
PB12N21
PB12N22
PB12N23
PB12N24
Drift Level(%)
COLLAPSE
MODERATE
Extensive large cracks.
Spalling of concrete in
weaker elements
SEVERE
Extensive crushing of
concrete.
Disclosure of buckled
reinforcements
MINOR
Minor cracks.
Partial crushing of
concrete columns
SLIGHT
Sporadic cracking
114 Chapter 4
 The damage distribution in the columns subjected to uniaxial and biaxial
loads was compared. As already observed by other authors, the plastic
hinge length was not significantly affected by the bidirectional loading.
Based on the test results, the length of the damaged zone in rectangular
columns was conditioned by the larger crosssection dimension.
 The drift limits for each damage state presented in FEMA 356 and VISION
2000 were compared with the drift demand associated with each damage
state observed in the tests. The drift limits proposed in the international
guidelines are in a good agreement with the damage observed in the biaxial
tests of RC columns.
 The initial stiffness in both directions is not significantly affected by the
biaxial loading path. However, biaxial loading tends to reduce the
maximum strength by approximately 8% in the column' strong direction
and 20% in the weak direction.
 The ultimate ductility is significantly lower in columns subjected to biaxial
loading paths when compared with uniaxial loading. This effect is more
pronounced in rectangular columns, with the ultimate ductility
approximately 50 to 75% lower in the weak direction and approximately
35% lower in the strong direction.
 The phase lags obtained for square columns are in agreement with the
results of other authors, as shown by the sheardrift curves and by the
relative rotation found between the horizontal load and displacement paths.
In rectangular columns, the results show that the path lag depends on the
section geometry and is lower when the force direction is close to the
column' strong direction.
 Comparison of the uniaxial and biaxial test results demonstrates a strong
correlation between the maximum and yielding strengths, i.e. the yielding
strength is approximately 70% of the corresponding maximum strength.
 There is almost no strength degradation during the first loading cycle for
each displacement level, but the degradation increases after a displacement
ductility factor of approximately 3 is reached. In the strength degradation
analysis, more pronounced strength degradation was observed in the biaxial
tests than in the corresponding uniaxial tests. The influence of the biaxial
loading path on the strength degradation is more significant for the
quadrangular path.
 Comparison of the ultimate drift capacity obtained from the test results on
the RC columns with the values given by the analytical expression of
Analysis of Experimental Test Results 115
EC83 showed that EC8 provides good results for uniaxial demands.
However, in columns where biaxial behaviour is expected, further
investigation is needed to evaluate the ultimate capacity.
 It was observed that biaxial loading can introduce higher energy dissipation
(circular, rhombus and cruciform load paths) than uniaxial loading, as
previously recognised by other authors. It was confirmed that the energy
dissipation also depends on the columns geometry. For a given imposed
maximum drift, among the load paths considered in this study, the circular
path was shown to be the most dissipative and the quadrangular load path
the less dissipative. The quadrangular load path dissipates even less energy
for a certain drift demand than the sum of the dissipated energy in two
independent unidirectional tests for the corresponding drift level.
 As stated by other authors, in the square columns tested with an axial load
stress of 2.33MPa, the total dissipated energy is independent of the loading
path. However for the tested square columns with higher axial load stress
(7.33MPa), it was verified that the total dissipated energy depends on the
load path.
 An expression relating the RC element' displacement ductility and
thenormalised dissipated energy was proposed for columns subjected to
uniaxial or biaxial loading and constant axial force, giving, for example, for
a displacement ductility of 4 leads to a corresponding normalised dissipated
energy of about 12.
 From the analysis of the viscous damping of each independent direction of
columns when tested biaxially, a larger dispersion of the damping in each
direction was found, however, it was verified that the viscous damping
strongly depends on the biaxial load path. The repetition of cycles, for the
same maximum displacement level, has almost no influence on the
equivalent damping.
 The equivalent biaxial damping was computed with the results of each
biaxial test, presenting a huge dispersion. The equivalent biaxial damping is
highly dependent on the load path.
 Two simplified expressions were proposed, based on the experimental
results, allowing a rough estimation of the equivalent biaxial damping in
RC columns subjected to biaxial loading. Although recognizing a possible
overestimation of the equivalent damping with the adopted calculation
method, these equations represent a first attempt for estimating the
equivalent damping of columns under biaxial loading. However, these
116 Chapter 4
expressions should be corrected based on results from nonlinear time
history analysis and/or dynamic shaking table tests.
 Based on the inherent concepts of the Park and Ang uniaxial damage
index, two new expressions were proposed to represent the damage
evolution in RC elements under biaxial loading. These proposals were
calibrated with the tested columns. The new proposals for the biaxial
damage index should be calibrated with test results on building structures
subjected to bidirectional loadings.
Finally, it is worth emphasising that many questions are still open in the field of
the biaxial behaviour of RC columns, especially those questions associated with the
response dependency on the load paths. Therefore, additional experimental research
should be conducted in this field, particularly considering that the cyclic tests
performed in this work by imposing smooth and regular displacement paths may
not be representative of the displacements actually imposed by earthquakes.
Nevertheless, the research reported here is expected to contribute for a better
understanding of the biaxial response of RC columns and for the calibration of
suitable numerical models for the biaxial lateral response of RC columns under
cyclic loading reversals.
Chapter 5
NonLinear Analysis of RC
Columns Subjected to Biaxial
Loads
5.1 Introduction and objectives
It is widely known that, in the analysis of structures subjected to seismic demands,
the use of numerical models based on nonlinear behaviour laws and hysteretic
rules allows for a more rigorous representation of the seismic response of the
structures [172].
In this framework, the present chapter focuses on the evaluation of different
nonlinear modelling strategies for predicting the response of reinforced concrete
columns subjected to biaxial bending, combined with constant axial load. To this
end, after being presented the results in Chapter 4 the tested columns were
simulated with different modelling strategies and the results obtained with each
model are discussed and compared in the present chapter. The three numerical
strategies under comparison are the distributed plasticity, force and displacement
based formulations, and the concentrated plasticity.
The main results of the numerical analyses described in this chapter, and its
principal conclusions, were reported in a paper submitted to the Earthquake
Engineering and Engineering Vibration journal [173].
118 Chapter 5
5.2 Numerical tool and modelling strategies
The numerical analyses developed and described in the present Chapter with
different nonlinear modelling strategies were performed using the computer
program SeismoStruct [23]. The program includes models for the representation of
the behaviour of spatial frames under static and/or dynamic loading, considering
both material and geometric nonlinearities. With this software, seven types of
analyses can be performed, namely: dynamic and static timehistory, conventional
and adaptive pushover, incremental dynamic analysis, modal analysis and static
analysis (possibly nonlinear) under quasipermanent loading.
The software allows for the use of elements with distributed inelasticity (force or
displacementbased formulations) and elements with lumpedplasticity (with fixed
length, i.e. the socalled plastichinge). Fibre discretization is adopted to represent
the behaviour at section level (see Figure 81), where each fibre is associated with a
uniaxial stressstrain law. The sectional momentcurvature state of the beam and
column elements is then obtained through the integration of the nonlinear uniaxial
stressstrain response of the individual fibres into which the section has been
subdivided.
Figure 81 Fibre based modelling (adapted from [174])
5.3 Element modelling strategies
In this work three nonlinear modelling strategies are compared, based on: elements
with lumpedplasticity (Figure 82a) and elements with distributed inelasticity, the
later with forcebased formulation (Figure 82b), and with displacement based
formulation (Figure 82c).
Reinforcement
steel
Unconfned
concrete
Confned
concrete
NonLinear Analysis of RC Columns Subject to Biaxial Loads 119
Figure 82 Modelling strategies with corresponding control section: a) lumpedplasticity
element; b) distributed inelasticity element with forcebased formulation; c) distributed
inelasticity element with displacementbased formulation
For the applications made with the different modelling strategies, decisions for each
specific modelling were taken based on the results of parametric studies performed
by other authors [18, 86, 175]. Different studies have proposed expressions to
estimate the plastic hinge length (Lp) of RC elements to be adopted in lumped
plasticity models [127]. Priestley and Park [176] proposed a formulation that
estimates the plastic hinge length based on the distance of the critical section from
the point of contraflexure and on the diameter of the longitudinal reinforcement
bars. Based on such work, Paulay and Priestley [52] reported that, for typical RC
columns, the plastic hinge length is approximately equal to half of the crosssection
depth. Therefore, in the simulations performed for the uniaxial tests, it was
considered half of the crosssection depth. For the biaxial tests, based on
experimental evidence, other authors have concluded that the plastic hinge length
is not strongly affected by biaxial loading [66]. The plastic hinge length in
rectangular columns under biaxial loading assumes approximately the length
observed in the column tested uniaxially in its strong direction [177]. In the
analyses performed in this study, half of the larger dimension of the crosssection
was considered as the plastic hinge length.
For the force based formulation, seven integration points were considered, based on
the results of Calabrese et al. [175]. These authors point out that at least six
integration sections are needed in order to obtain a completely stabilized prediction
of the local response.
According to the results of Calabrese et al. [175], for displacement based
formulations a good approximation to a cantilever column response can be obtained
with a mesh discretization of at least four elements, with two GaussLegendre
points per element, if all elements have the same length. Considering this, and
taking into account the concentration of the nonlinear response close to the fixed
Node
Integration point
0
.
4
6
9
*
L
/
2
0
.
8
3
0
*
L
/
2
La
La
Lp
2
Lp
2
Lp
La = L  2.Lp
3
L
Lp
Lp
L
i n
e
a
r
e
l a
s
t
i c
120 Chapter 5
end of the column (plastic hinge length), a discretization was adopted with six
elements having the sizes showed in Figure 82.
5.4 Materials properties
The consideration of nonlinear material behaviour for the prediction of the RC
columns response requires accurate modelling of the uniaxial material stressstrain
cyclic response. Thus, the present section describes the adopted constitutive model
for the concrete and reinforcing steel used herein and the values considered for each
parameter.
5.4.1 Concrete
For the concrete, it is used a uniaxial model based on the Madas [30] proposal,
which follows the constitutive law proposed by Mander et al. [31]. The cyclic rules
included in the model for the confined and unconfined concrete were proposed by
MartinezRueda and Elnashai [32]. The confinement effects provided by the
transverse reinforcement were considered through the rules proposed by
Mander et al. [31], whereby constant confining pressure is assumed throughout the
entire stressstrain range, traduced by the increase in the peak value of the
compression strength and the stiffness of the unloading branch [23].
Table 23 presents the properties considered in the numerical models. Material
mechanical parameters were defined based on the samples test results presented in
Chapter 3.
Table 23 Concrete mechanical parameters for the numerical model
Compressive strength
f
c
(MPa)
Tensile strength
f
t
(MPa)
Strain at peak strength
c
(%)
Confinement
factor
*
N01N04 39.4 3.94 0.36 1.11
N05N08 21.4 2.35 0.34 1.12
N09N12 24.4 2.57 0.36 1.12
N13N16 21.7 2.98 0.33 1.06
N17N18 36.3 3.24 0.36 1.12
N19N22 43.14 3.28 0.42 1.12
N23N24 36.3 3.24 0.36 1.06
*
The confinement factor was obtained with the Mander et al. [178] proposal
5.4.2 Reinforcement steel
The uniaxial model proposed by Menegotto and Pinto [33], coupled with the
isotropic hardening rules proposed by Filippou et al. [34], was adopted for the steel
reinforcement representation in these analyses. This steel model does not represent
NonLinear Analysis of RC Columns Subject to Biaxial Loads 121
the yielding plateau characteristic of the mild steel virgin curve. The model takes
into account the Bauschinger effect, which is relevant for the representation of the
columns stiffness degradation under cyclic loading. The input parameters of the
model are: the yield strength (f
y
); the elastic Young modulus (E
s
); the strain
hardening ratio (r) and five parameters to describe the transition from elastic to
plastic branches (R
0
, a
1
, a
2
, a
3
, and a
4
). The considered model parameters are
summarized in Table 24. All the adopted values are in accordance with the
properties obtained in the material test samples presented in Chapter 3.
Table 24 Steel mechanical parameters for the numerical model
Elasticity
modulus
Yield
strength
Strain
hardening
parameter
Transition
curve initial
shape
Transition
curve shape
Isotropic
hardening
E
s
(GPa) f
y
(MPa) r () R
0
a
1
a
2
a
3
a
4
N01N04 194.7 432.63 2.71
20 18.5 0.15 0.025 2 N05N16 203.46 429.69 2.69
N17N24 189.53 450.26 3.32
5.5 Comparison between modelling strategies
The three modelling strategies were used to reproduce the response of the
specimens tested. As stated previously, material mechanical parameters were
defined based on the results of material tested samples. For the parameters
controlling the cyclic response of the reinforcement steel the default values were
adopted as indicated in the software manual [23]. For each analysed column, the
axial load and the corresponding horizontal displacement law were imposed in
accordance with the performed test. The numerical results for each modelling
strategy were compared with the experimental results and are discussed in the next
sections.
5.5.1 Shear drift envelopes
The experimental sheardrift envelope curve obtained for each column is compared
with the results of the nonlinear models in terms of initial stiffness and in terms of
evolutions of tangent and secant stiffness. As an example, Figure 83 shows the
sheardrift envelopes in one direction for the columns with crosssectional
dimensions of 30x50cm
2
for uniaxial and biaxial tests (columns N09 to N12 and
N18). From the analysis of the sheardrift envelopes, good agreement was found in
the numerical representation of the experimentally measured response. The
differences found between the different modelling strategies are discussed next.
122 Chapter 5
Figure 83 Sheardrift envelopes for 30x50cm
2
columns (measured and calculated)
5.5.1.1 Global comparison
In order to compare the envelopes obtained with the experimental results the
correlation coefficient R
2
was used. The R
2
is a well known statistic measure that
0 1 2 3 4
0
25
50
75
100
125
150
175
200
S
h
e
a
r
X
(
k
N
)
Drift X (%)
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
PB01N09
X
Y
X
Y
0 1 2 3 4 5 6
0
25
50
75
100
125
150
175
200
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
PB02N10
X
Y
X
Y
0 1 2 3 4
0
25
50
75
100
125
150
175
200
S
h
e
a
r
X
(
k
N
)
Drift Y (%)
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
PB12N11
X
Y
X
Y
0 1 2 3 4
0
25
50
75
100
125
150
175
200
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
X
Y
X
Y
PB12N11
0 1 2 3 4
0
25
50
75
100
125
150
175
200
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N12
X
Y
X
Y
0 1 2 3 4
0
25
50
75
100
125
150
175
200
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
X
Y
PB12N12
X
Y
0 1 2 3 4
0
25
50
75
100
125
150
175
200
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N18
X
Y
X
0 1 2 3 4
0
25
50
75
100
125
150
175
200
Experimental
Numerical  FB
Numerical  DB
Numerical  LP
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N18
X
Y
X
NonLinear Analysis of RC Columns Subject to Biaxial Loads 123
provides information about the goodness of fit, in which a R
2
=1.0 represents a
perfect correlation between the numerical and experimental data. In this work it
was assumed that 0.75 is the minimum R
2
value for a good fitting. Therefore, based
on the obtained results as presented in the circular charts illustrated in Figure 84,
the following can be observed:
 All the modelling strategies exhibit similar correlation coefficients, larger
than 0.75 in most of the cases, which mean successful representations of the
envelopes.
 The lumped plasticity element shows the lowest correlation factors.
 The correlation coefficients results are similar for the strong (X) and weak
(Y) columns directions.
a)
b) c)
Figure 84 Correlation coefficients (R
2
) for the sheardrift envelopes between experimental
and numerical results: a) uniaxial texts; b) biaxial tests strong direction (X); c) biaxial
tests weak direction (Y)
5.5.1.2 Initial stiffness comparison
The column initial stiffness values calculated from the sheardrift envelopes curves
obtained with the three numerical models, were compared with the corresponding
initial stiffness derived from the experimental tests.
Distributed Inelasticity  DB
Distributed Inelasticity  FB
Lumped Plasticity
124 Chapter 5
Figure 85 shows circular charts of the ratio between the experimental and
numerical initial stiffness obtained for all columns. From these plots analysis, the
following can be concluded:
 For the uniaxial tests the ratio assumes values close to 1, indicating that all
models represent well the columns initial stiffness, with a maximum
underestimation of about 25%.
 For the biaxial results, a larger dispersion was found in the calculated
initial stiffness, with a underestimation in the column strong direction
(around 2550%) and an overestimation in the weak direction
around 25%).
a)
b)
c)
Figure 85 Initial stiffness ratio between experimental and numerical results: a) uniaxial
texts; b) biaxial tests strong direction (X); c) biaxial tests weak direction (Y)
5.5.1.3 Secant stiffness comparison
The evolution of the columns secant stiffness with the drift demand was evaluated
by comparing the peaktopeak secant stiffness of the first cycle for each drift
demand level. Figure 86 shows examples of the secant stiffness reduction with the
drift demand for the columns PB01N1, PB02N6, PB12N7 and PB12N12. The
secant stiffness evolutions for all the tested columns were compared with the
corresponding values obtained with the developed numerical models. Regardless of
Distributed Inelasticity  DB
Distributed Inelasticity  FB
Lumped Plasticity
NonLinear Analysis of RC Columns Subject to Biaxial Loads 125
the displacement history (uniaxial or any path for biaxial loading), the secant
stiffness evolution obtained from the tests is quite well captured with all the
modelling strategies.
Figure 86 Secant stiffness evolution for columns PB01N1, PB02N6, PB12N7 and PB12
N12: experimental and numerical results
5.5.1.4 Tangent stiffness comparison
The variation of the tangent stiffness determined from the sheardrift envelope
curves was analysed for all the studied columns. Figure 87 shows examples of the of
tangent stiffness variations (experimental and numerical) with the drift demands,
namely for columns PB01N1, PB02N6, PB12N7 and PB12N12. From the results
obtained for all columns, the following was observed:
 The tangent stiffness presents generally similar evolutions trends for the
three modelling strategies.
0 1 2 3
0
5000
10000
15000
20000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB01N01
X
Y
0 1 2 3 4 5
0
5000
10000
15000
20000
25000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
X
Y
PB02N06
0 1 2 3 4
0
5000
10000
15000
20000
25000
30000
35000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N07
X
Y
0 1 2 3 4
0
5000
10000
15000
20000
25000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N07
X
Y
0 1 2 3
0
15000
30000
45000
60000
75000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N12
X
Y
X
Y
0 1 2 3
0
5000
10000
15000
20000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N12
X
Y
X
Y
126 Chapter 5
 Compared with the experimental results, numerical models show difficulties
in representing the tangent stiffness for drift demands (less than 1%). This
fact was already reported in the analysis of the differences obtained for the
initial stiffness. For drift demands larger than 1%, a good representation of
the tangent stiffness evolution is obtained with the numerical models,
corresponding to the postyielding plateau zone (approximately zero
stiffness).
 However, for drift demands larger than those corresponding to the
beginning of the column strength degradation, the numerical models do not
accurately represent the tangent stiffness evolution. This aspect is more
evident for the columns under biaxial loading conditions, since the strength
degradation starts for lower drift demands, as can be observed in the
examples presented in Figure 87, where the postyielding plateau is shorter.
Figure 87 Tangent stiffness evolution for columns PB01N1, PB02N6, PB12N7 and
PB12N12: experimental and numerical results
0 1 2 3
5000
0
5000
10000
15000
20000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB01N01
X
Y
0 1 2 3 4 5
10000
5000
0
5000
10000
15000
20000
25000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
X
Y
PB02N06
0 1 2 3 4
15000
10000
5000
0
5000
10000
15000
20000
25000
30000
35000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N07
X
Y
0 1 2 3 4
10000
5000
0
5000
10000
15000
20000
25000
30000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N07
X
Y
0 1 2 3
30000
15000
0
15000
30000
45000
60000
75000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift X (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N12
X
Y
X
Y
0 1 2 3
5000
0
5000
10000
15000
20000
S
t
i
f
n
e
s
s
(
k
N
/
m
)
Drift Y (%)
Experimental
Distributed Inelasticity  FB
Distributed Inelasticity  DB
Lumped Plasticity
PB12N12
X
Y
X
Y
NonLinear Analysis of RC Columns Subject to Biaxial Loads 127
5.5.2 Cyclic response
This section focused on the assessment of the models accuracy for representing the
cyclic response of the RC columns under study. The whole set of obtained shear
drift diagrams are fully presented in Appendix B. In this section only the results
for columns PB01N1, PB02N6, PB12N7 and PB12N12 are presented in order to
compare experimental and numerical results.
5.5.2.1 Overall analysis
The general analysis of the whole cyclic columns response obtained with the three
modelling strategies and the comparison with the corresponding experimental
results allow concluding that:
 Even when a good representation of the global columns response is
achieved, for the largest deformation demands the numerical response may
not able to capture well the strength degradation, which is associated to the
buckling of the reinforcing steel bars.
 In many cases, the unloadingreloading phase of the columns cyclic
response obtained with the numerical models does not accurately capture
the pinching effect observed experimentally. This can be justified by the
limitations of the models in the representation of the longitudinal
reinforcing steel slippage.
128 Chapter 5
Lumpedplasticity
Forcebased
Displacementbased
N
0
1
N
0
6
N
0
7
X
N
0
7
Y
N
1
2
X
N
1
2
Y
Figure 88 Sheardrift response for columns PB01N1, PB02N6, PB12N7 and PB12N12:
experimental and numerical results
L
Lp
Lp
L i n e a r e l a s t i c
0.469*L/2
0.830*L/2
La
La
Lp
2
Lp
2
Lp
La = L  2.Lp
3
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
NonLinear Analysis of RC Columns Subject to Biaxial Loads 129
5.5.2.2 Shear force evolution
In order to evaluate the accuracy of the numerical models adopted for representing
the shear force evolution experimentally obtained, the frequency domain error
(FDE) index [35, 36] was calculated. The FDE index [36] measures the deviation
between two waveforms, in this case the shear force measured in the test and the
corresponding calculated shear force for each modelling strategy, which are
obtained for the same measured displacement (i.e. for the same time instant). The
FDE index quantifies amplitude and phase deviations between two signals, giving
an error factor with a value between 0 and 1. In this analysis, it is considered that
a FDE index larger than 0.75 represents a poor correlation and a value below 0.25
represents a very good correlation. These limits were proposed in [35].
The FDE indices obtained for all analyses are plotted in the circular charts
represented in Figure 89. From the results analysis, the following comments can be
drawn:
 For all studied columns, under uniaxial and biaxial loading, the three
numerical models give good estimates of the shear force evolutions recorded
in the experimental tests, because the calculated FDE index was just
slightly larger than 0.25. Moreover, no significant differences were found in
the FDE index for the three modelling strategies used for each tested
column.
 For all columns subjected to biaxial demands, the FDE index is lower in
the weak direction (Y) than in the strong direction (X), which is in
accordance with the larger differences identified for the latter direction in
terms of stiffness evolution and strength degradation.
130
Figu
force
5.5.2
The
dissip
with
the e
of th
energ
ure 89 FDE
e evolutions:
2.3 Cumu
accuracy of
pated energ
the values
evolution of
he ratio betw
gy.
E index obtain
a) uniaxial t
ulative ener
f the nonli
gy by comp
obtained fr
the cumula
ween the nu
a)
b)
ned from the
tests; b) biax
weak d
rgy dissipa
near models
paring the v
rom the test
ative dissipa
umerical and
)
)
comparison
xial tests str
direction (Y)
ation
s is charact
values calcu
ts on colum
ated energy
d experimen
of numerical
rong direction
)
terized here
ulated via
mns. Figure
and Figure
ntal values
Distribute
Distribute
Lumped P
and experim
n (X); c) biax
in terms o
the numeri
90 shows e
91 includes
of the total
ed Inelasticity 
ed Inelasticity  F
Plasticity
Chapter 5
c)
mental shear
axial tests
of the total
ical models
examples of
s the charts
l dissipated
DB
FB
NonLinear Analysis of RC Columns Subject to Biaxial Loads 131
Figure 90 Evolution of the cumulative dissipated energy for columns PB01N1, PB02N6,
PB12N7 and PB12N12: Experimental and numerical results
Based on the obtained results, the following aspects can be concluded:
 The main differences observed in the cumulative dissipated energy
correspond to an underestimation of the dissipated energy for drift
demands, associated with a quasilinear numerical response, and to an
overestimation for large demands associated with the limitations of the
numerical models in representing of the strength degradation and pinching
effect (see Figure 90).
 For some columns analysed under uniaxial loading conditions, considerable
overestimations of the total dissipated energy were obtained with the
numerical models (Figure 91a). This effect was not observed in any column
under biaxial loading. The overestimation of the dissipated energy is
justified by the inadequacy of the numerical models in representing the
strength degradation.
0
10
20
30
40
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
Step
PB01N01
X
Y
0
50
100
150
200
250
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
Step
PB02N06
X
Y
0
25
50
75
100
125
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
X
D
i
r
e
c
t
i
o
n
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
Step
PB12N07
X
Y
0
25
50
75
100
125
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
Y
D
i
r
e
c
t
i
o
n
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
PB12N07
X
Y
Step
0
25
50
75
100
125
150
175
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
X
D
i
r
e
c
t
i
o
n
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
PB12N12
X
Y
X
Y
Step
0
25
50
75
100
125
150
175
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
Y
D
i
r
e
c
t
i
o
n
(
k
N
.
m
)
Experimental
Distributed inelasticity  DB
Distributed inelasticity  FB
Lumped Plasticity
PB12N12
X
Y
X
Y
Step
132 Chapter 5
 Results for the columns subjected to biaxial loading, analysed for each
direction independently, show variations in the total dissipated energy
between 25 and 50% (Figure 91c and Figure 91d), which are larger for the
weak direction (Y).
 Generally, the sum of the total dissipated energy in each direction exhibits
an overestimations with the adopted numerical strategies (Figure 91b). For
columns N22, N23, and N24, the larger axial load level anticipates the
collapse for lower drift demands and therefore, the total dissipated energy
calculated from the numerical response underestimates the experimental
value. This fact is justified by the less adequacy of the models in
representing well the energy dissipation for lower deformations, associated
with a quasilinear numerical response in this stage.
 For each studied column under biaxial loading conditions, similar ratios
were obtained with the three modelling strategies. However, for the weak
direction the distributed plasticity model with forcebased formulation
presents higher differences in the total dissipated energy. Finally the
comparison of the sum of the total dissipated energy in both directions
shows that the lumped plasticity modelling strategy presents larger
deviations.
NonLinear Analysis of RC Columns Subject to Biaxial Loads 133
a)
b)
c) d)
Figure 91 Ratio between the total dissipated energy obtained with the numerical models
and with the experimental value: a) uniaxial tests; b) biaxial tests total (X + Y); c)
biaxial tests strong direction (X); d) biaxial tests weak direction (Y)
5.6 Final remarks
In this Chapter, three modelling strategies were assessed concerning their adequacy
for representing RC columns responses, under uniaxial and biaxial loading
conditions. From the analysis of the obtained results, similar levels of adequacy
were observed either using the distributed inelasticity (force and displacement
formulations) or the lumped plasticity modelling strategies. However, based on the
previously made comparisons some conclusions can be drawn:
 The initial stiffness obtained with all the modelling strategies for the
simulation of the biaxial tests presents differences, in the range of 25 to
50%.when compared with the experimental values.
 The evolution of the columns secant stiffness was well represented with all
modelling strategies. However, considerable differences were observed, for
all modelling strategies, between the numerical and experimental tangent
stiffness evolutions, mainly for the lower drift demands and in the last
Distributed Inelasticity  DB
Distributed Inelasticity  FB
Lumped Plasticity
134 Chapter 5
phase of the columns responses. This aspect confirms the difficulties
normally found in representing (numerically) the response of columns with
models based on incremental procedures, dependent of the tangent stiffness.
 The cyclic response obtained with the three modelling strategies for all
columns was found to be satisfactory, but difficulties were found in
capturing the strength degradation for larger drift demands. Also, most of
the models show limitations in representing the pinching effect in the
unloadingreloading stage.
 All modelling strategies exhibit difficulties in the simulation of the energy
dissipation evolution. The difference in the total dissipated energy was
found to range between 25 and 50%.
Many questions are still open in the modelling of the biaxial response of RC
columns. The considered modelling strategies can accurately predict the cyclic
response of the columns until the strength degradation begins. For uniaxial loading,
this aspect is not so relevant, since strength degradation starts for larger drift
demands in comparison with biaxial loadings.
All the adopted modelling strategies show similar deviations to the columns
response obtained in the experimental tests. This fact points towards the use of a
lumped plasticity modelling strategy for the representation of RC elements, since
the strategy is simple and the element discretization is only dependent on the
plastic hinge length, which has a direct physical interpretation. Furthermore, when
used for the analysis of complex RC building structures, the lumped plasticity
modelling strategies lead to more reduced computational time and the analysis is
concentrated in the element critical sections.
Chapter 6
Simplified Model for the Non
Linear Behaviour of RC
Columns under Biaxial
Bending
6.1 Introduction
Present seismic design recommendations consider that buildings should respond
elastically to small magnitude earthquakes, although allowing for nonlinear
response for larger magnitudes; therefore, adequate nonlinear models to account
for the hysteretic behaviour of RC structures are a present need.
As emphasized in Chapter 2, the study of the tridimensional response of RC
buildings is an important subject, associated to earthquake actions or to building
irregularities, which induces a biaxial bending demands combined with axial load in
the columns. Nowadays, even with the different modelling strategies available for
the simulation of biaxial cyclic behaviour of RC elements with axial force, its
applicability in current practice as well as in codes and standards is yet not
accepted as it is for the uniaxial simplified models.
In this framework the present chapter focuses on the proposal of an upgraded
simplified model for the representation of the biaxial bending response of columns
with axial force, which is based in existing uniaxial hysteretic CostaCosta
model [179].
136 Chapter 6
The general formulation of the proposed model is established by analogy and
comparison with the biaxial formulation of the BoucWen smooth hysteretic
model [9]. Although retaining most of the physical meaning embodied in the Bouc
Wen model, the structural modelling strategy adopted keeps the simplicity and
versatility of the original piecewise linear (PWL) numerical tool.
The model formulation is presented in a general forcedisplacement form to be in
agreement with the experimental results selected to test the model performance.
Moreover, the specific biaxial bending response of columns is easily obtained by
expressing the forcedisplacement law in terms of momentcurvature.
The simplified numerical model described in this chapter was summarily presented
in a paper submitted for eventual publication in the journal Engineering
Structures [180].
6.2 Uniaxial hysteretic model
For the development and validation of the simplified biaxial model, the Costa
Costa uniaxial hysteretic model [32, 179] was adopted, which is briefly described in
the next paragraph.
This model represents a generalisation of the original Takeda model [32, 181] with
a trilinear skeleton curve for monotonic loading, defined by the cracking point
(d
c
;F
c
) and the yielding point (d
y
;F
y
), which includes pinching, stiffness degradation
and strength degradation effects.
Unloadingreloading loops prior to yielding in either direction are bilinear, with
slopes equal to those of the precracking and postcracking branches in the virgin
loading branch. After yielding, pinching is modelled by a bilinear reloading curve.
The first branch of the reloading stage has an inferior slope (see Figure 92  branch
8), while the second branch heads to the most extreme point of any previous post
yield excursion in the direction of the reloading. For this propose, the stiffness K
s
is
multiplied by the factor (d
y
/d
m
)

, where d
y
represents the yield displacement, d
m
the
maximum response displacement and  is a positive constant. Thus:
( )

m y
0 m
m
s
d / d
d d
F
K
=
(36)
where F
m
is the force at the previous maximum response point and d
0
is the
deformation at the load reversal point.
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 137
Figure 92 Pinching effect in the CostaCosta [179] model
The unloading stiffness K
d
, after yielding, is reduced relative to the cracked secant
stiffness K
e
by the factor (d
y
/d
m
)
o
, where o is a positive constant (see Figure 93).
Figure 93 Unloading stiffness in the CostaCosta [179] model
Postyielding strength and stiffness degradation with cycling loading is modelled by
directing the reloading branch, after modification for pinching, towards a point at a
displacement equal to d
m
and at a force F
m
=(1)F
m
, where is the Wang and
Shah damage index [182] (see Figure 94). After reaching this end point of the
reloading branch, further loading takes place parallel to the postyielding stiffness
of the virgin loading curve.
6
F

m
m
d

12
13
7
5
8
K
1
2
9
F
4
d
m
+
d
3
11
10
m
F
+
s
d d
c y
F
c
F
y
F
d
m
d
Kd
e
K
138 Chapter 6
Figure 94 Strength degradation in the CostaCosta [179] model
6.3 Biaxial BoucWen model
6.3.1 Background
The original formulation of the BoucWen model, cast within the endochronic
theory framework was presented by Bouc [183], for uniaxial behaviour
representation (in terms of ForceDisplacement, Fu), and it was later generalized
by Wen [95]. The generalized model expresses the restoring force as a combination
of an elastic force and a plastic force:
Z
y
F u K + = ) 1 ( F o o
(37)
where K is the initial stiffness, o the postyielding stiffness ratio, F
y
the yielding
force, and Z is a hysteretic parameter given by:
y
u
n
Z u
n
Z Z u u A
Z

1
(38)
in which A, ,  and n are dimensionless parameters controlling the shape and
magnitude of the hysteresis loops and u
y
is the yielding displacement of the system.
As usual in endochronic model types the incremental form of Equation (37) can be
written as:
Z
y
F u K F
+ = ) 1 ( o o
(39)
where each parameter rate (JF, Ju, JZ) corresponds to inscrements, which are often
denoted also by F
, u , Z
.
m
d
m
m
F
F
F
d
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 139
Therefore, Equation (39) can be joined with Equation (38) into a single expression
which defines the incremental restoring force
i
F
\

+ =
n
i
Z
i
u
n
i
Z
i
Z
i
u
i
u A K
i
u K
i
F
 o o
1
1
(40)
The variable Z
i
can be eliminated from Equation (40) by rewriting Equation (37) in
the form:
y
i
F
u K
i
F
i
Z
=
) 1 ( o
o
(41)
considering that the global restoring force F
i
results from:
i
F
i
F
i
F
+
=
1
(42)
where F
i1
represents the global restoring force for the previous displacement
increment.
Later, this uniaxial formulation was extended by Park, Wen and Ang [95, 111] to
define a biaxial forcedeformation model with coupled differential equations. This
model was then used and modified by Kunnath and Reinhorn [96] to model the
behaviour of RC columns under biaxial loads. Later on, the model was generalised
by Casciati [97] and also by Wang and Wen [98], which resulted in two different
formulations of the initial biaxial model. Since the Wang and Wen [98] formulation
is simpler, it was selected for to implementing the biaxial model proposed in this
work. Nevertheless, the same mathematical reasoning can be applied to the
Casciati [97] form.
The biaxial construction of the BoucWen model in the Wang and Wen [98] form
follows the same general idea as for the uniaxial case. The restoring forces for both
directions are defined by:
(43)
in which the involved parameters have the same meaning as for the uniaxial case,
Equation (37), but are now referred to the two orthogonal directions X and Y, by
the subscripts x and y respectively. The hysteretic parameters Z
x
and Z
y
are then
defined by the following coupled differential equations, where all the parameters
( )
( )
+ =
+ =
y
y
y y y y y y
x
y
x x x x x x
Z F 1 u K F
Z F 1 u K F
o o
o o
140 Chapter 6
involved have also the same meaning as for the uniaxial case and sign() refers to
the mathematical signum function.
(44)
As for the case of the uniaxial model, Equation (43) can also be reformulated into
an incremental form:
(45)
which joined with Equation (45), allows defining the following system of coupled
equations expressing the incremental restoring orthogonal forces
x
F
and
y
F
:
(46)
As for the uniaxial case, the variables
i
x
Z and
i
y
Z may be eliminated from
Equation (46) by rewriting Equation (43) in the form:
(47)
considering that the global restoring forces
i
x
F and
i
y
F result from:
(48)
( )   ( )  
( )   ( )  
+ +
=
+ +
=
y
y
x x
1 n
x y x y y
1 n
y y y y
y
y
x
y y
1 n
y x y x x
1 n
x x x x
x
u
Z u sign Z Z u Z u sign Z Z u u A
Z
u
Z u sign Z Z u Z u sign Z Z u u A
Z
 
 
( )
( )
+ =
+ =
y
y
y y y y y y
x
y
x x x x x x
Z F 1 u K F
Z F 1 u K F
o o
o o
( ) ( )  
( )  
( ) ( )  
( )  
(
+
+ + =
(
(
+
+ + =
i i i i i
i i i i i i i i
i i i i i
i i i i i i i i
x x
1 n
x y x
y y
1 n
y y y y y y y y y y
y y
1 n
y x y
x x
1 n
x x x x x x x x x x
Z u sign Z Z u
Z u sign Z Z u u A K 1 u K F
Z u sign Z Z u
Z u sign Z Z u u A K 1 u K F

 o o

 o o
( )
( )
=
=
y
y y
y y y y
y
y
x x
x x x x
x
F 1
u K F
Z
F 1
u K F
Z
i i
i
i i
i
o
o
o
o
+ =
+ =
i 1 i i
i 1 i i
y y y
x x x
F F F
F F F
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 141
6.3.2 Framework of the proposed biaxial model
Considering the definition of the incremental orthogonal forces
i x
F
and
i
y
F
given
by Equation (46), the first part of this system can be written, by a simple
mathematical transformation, as:
(49)
which matches Equation (40) applied to the X and Y directions, thus
corresponding to the uniaxial incremental restoring forces calculated for each
direction without biaxial interaction. In order to simplify the notation, these
incremental forces will be denoted as
i x uni
F
and
i
y uni
F
, respectively. The
remaining part of the system corresponds to the correction factors
i
fx
C and
i
fy
C
accounting for the interaction between the two loading directions:
(50)
Therefore, based on Equations (49) and (50), a condensed form of Equation (45)
can be written as:
=
+
=
fyi
C
i
y uni
F
i
y
F
fxi
C
i
x uni
F
i
x
F
(51)
Considering that the incremental forces
i x uni
F
and
i
y uni
F
and
i
y uni
F

.

\

+ + =

.

\

+ + =
i i i i i i i i
i i i i i i i i
y y
1 n
y y y y y y y y y y
x x
1 n
x x x x x x x x x x
Z u sign Z Z u u A K 1 u K F
Z u sign Z Z u u A K 1 u K F
 o o
 o o
( ) ( )  
( ) ( )  
+ =
+ =
i i i i i i
i i i i i i
x x
1 n
x y x y y fy
y y
1 n
y x y x x fx
Z u sign Z Z u K 1 C
Z u sign Z Z u K 1 C
 o
 o
142 Chapter 6
additional parameter o, which was included in Equation (51) in order to scale the
level of interaction between the two loading directions. The final formulation of the
proposed method is then defined as written in Equation (52), which states that for
each incremental displacements vector ( )
i
y i x
u u ; , the incremental forces
i x uni
F
and
i
y uni
F
+ =
+ =
i i i
i i i
fy y uni y
fx x uni x
C F F
C F F
o
o
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 143
inspired algorithms. The immune algorithms and the methods based in neural
networks and artificial intelligence are also often used.
In this work, aiming at calibrating the parameters for the hysteretic biaxial model
based on uniaxial models with an interaction function, a gradient basedmethod
(LevenbergMarquardt (LM) method) was adopted, sequentially associated with an
evolutionary algorithm method (real search space EA), grouped. With these global
cascade algorithms, it is intended to aggregate the advantages of both algorithms
and to minimise their disadvantages [186].
Figure 95 Inverse problem scheme
The nonlinear optimization methods based on the gradient function can find local
minima through an iterative process for any initial solution. However, frequently
these results do not correspond to the absolute minimum. In most of the situations,
the solution depends on the initial estimation of parameters, thus leading to non
objective results. Tin order to reduce the dependence of the initial solution, an all
inclusive set of experimental results can be selected for discriminating the different
solutions obtained.
Evolutionary algorithms (EA), which represent genetic algorithms (GA),
evolutionary programming and strategies and their recombined algorithms, are
robust optimization techniques. Evolutionary algorithms are search and
optimization techniques proposed by John Holland, inspired in the process of
evolution through natural selection suggested by Darwin. These algorithms are
often used nowadays, leading to good results and performance for the solution of
structural problems, the characterisation of constitutive models, etc. [184].
For the application of these models, the SDL/SiDoLo optimisation lab computer
program was used [187]. The program was designed for specific engineering inverse
problems, such as parameter identification and initial shape optimization problems.
It inherits the wealth of experience gained in such problems by the previous
SIDOLO code, and adds the latest developments in direct search optimization
algorithms [184]. The SDL framework can be used, controlling every step of the
engineering optimization procedure, as well as aiding users that are not familiar
Simulation
Software
Optimization
Software
Input data
Uniaxialskeletoncurves
Hystereticparameters
Initialinteractionfunction
parameters
Output data
Shearforces
ProcessEvolutionproperties
Comparison
Newinteractionfunction
parameters
Comparison
Numericallyobtainedresults
Vs
Experimentaldata
144
with
an op
6.4.
The
prese
prese
most
type
calib
inter
the r
6.4.2
In or
inter
a set
RC c
ratio
was c
for d
formu
(0 a
analy
An i
colum
inter
gradi
numb
and
Reinh
programmi
ptimization
.2 Optim
calibration
ented in the
ented in Sec
suitable fo
selection,
ration of t
action scalin
results for a
2.1 Predic
rder to defin
action, as a
t of numeric
columns we
and axial
considered.
different d
ulation and
and 90) an
yses were pe
interaction
mn, based
action func
ientbased
ber of varia
i
fy
C were
horn [8] we
Figure 96
ng language
problem for
mization
n of the in
e previous se
ction 6.4.2.1
or the inter
the second
the parame
ng factor. T
group of nu
ction of the
ne the inter
a function de
cal analyses
ere defined,
load ratio (
The respon
irections C
considering
d five biaxi
erformed usi
scaling fact
on the Co
tion previou
optimization
ables, for the
assumed
re adopted
Column cro
es that wan
rmulation [1
of the in
nteraction f
ection, was
, it was inte
action scali
phase (pre
eters of th
The paramet
umerical ana
he scaling fa
action scalin
ependent of
was perfor
varying in
(see Table 2
nse of each c
Columns we
g a fibre dis
ial (at 11.25
ing the com
tor was det
ostaCosta u
usly present
n algorithm
e shape fact
equal to
( 0 = = 
osssection an
nt to use an
187].
nteractio
function, us
performed
ended to sel
ing factor (
esented in
he interacti
ters calibra
alyses.
factor equa
ng factor (o
f the section
rmed. To th
n crosssect
25). For the
column was
ere modelle
scretization
5, 22.5, 45
mputer progr
termined fo
uniaxial mo
ted in Equa
m was used
tors (,  a
the value
5 . and = n
nd definition
intuitive gr
on funct
sing the o
in two phas
lect the ana
o) function
Section 6.4
on function
tion corresp
ation
o) which cha
n properties
his aim, twe
tion dimens
e column, a
s obtained w
ed with a
at the secti
5, 67.5, 78
ram SeismoS
or each bia
odel couple
ation (52).
d [184]. In
nd n) neces
es suggeste
2 ).
of the lateral
raphical inte
ion
ptimization
ses. In the f
lytical expre
. After the
4.2.2), consi
n as well
ponds to the
aracterizes t
and loading
enty seven r
ions, reinfo
a cantilever
with pushov
forcebase
on level. Tw
8.75 angles
Struct [188].
axial respon
ed with the
For these a
order to r
ssary to calc
d by Kun
l loading dire
Chapter 6
erface with
n strategies
first phase,
ressiontype
expression
ists in the
as of the
e bestfit of
the level of
g direction,
rectangular
orcing steel
1.5m high
ver analysis
ed element
wo uniaxial
s) pushover
.
nse of each
e proposed
analyses, a
reduce the
culate
i
fx
C
nnath and
ection
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 145
Table 25 Crosssection properties
Model
Reference
Crosssection
dimensions [cmxcm]
Reinforcing steel
[% cross sec. area]
Axial load
ratio
M1
30x30
1.0%
0.1
M2 0.2
M3 0.3
M4
1.5%
0.1
M5 0.2
M6 0.3
M7
2.0%
0.1
M8 0.2
M9 0.3
M10
30x50
1.0%
0.1
M11 0.2
M12 0.3
M13
1.5%
0.1
M14 0.2
M15 0.3
M16
2.0%
0.1
M17 0.2
M18 0.3
M19
30x60
1.0%
0.1
M20 0.2
M21 0.3
M22
1.5%
0.1
M23 0.2
M24 0.3
M25
2.0%
0.1
M26 0.2
M27 0.3
Aiming at evaluating the goodnessoffit of the given interaction scaling factor, for
each column and for each direction calculation are made for, the difference between
the simulated response with the simplified (sim) model (with the interaction
function parameters) and with the refined numerical model (reference values, ref).
These differences are evaluated for each direction (X or Y) by the Relative Global
Error (RGE
direction
), as given in Equation (53). The combination of the error for the
two directions (RGE
total
) is calculated as presented in Equation (54).
R0E
dccton
%] = 1uu _1 
A
sm
A
c]
_
c
=1
(53)
R0E
totuI
= _ R0E
X
2
+R0E
(54)
146 Chapter 6
where A
sm
and A
c]
are, respectively, the simulated and reference values of the
potential energy associated with each pushover response.
After obtained the optimal values of the scaling factor (o) for each analysis (each
column and each direction), an expression type was selected (see Equation (55)),
which depends on the column properties, namely the crosssection dimensions (h
and b), the axial load ratio (); and loading direction ().
o = C
1
_
h
b
]
C
2
e
C
3
v
(tan)
C
4
(55)
The four constants (C
1
, C
2
, C
3
and C
4
) are to be obtained by optimization for all
pushover curves, as will be shown in the following section.
6.4.2.2 Global parameter identification
With the adopted equation for the scaling factor, the interaction function
parameters were optimized for all biaxial pushover curves, using a cascade
optimization strategy.
At this stage, the LevenbergMarquardt (LM) gradientbased method and an
evolutionary method (real search space EA) were grouped in sequential/cascade
strategies. Thus, as mentioned before, in order to combine the advantages of both
algorithms and minimise their disadvantages, the following sequence LM+EA+LM
was used. An important aspect in a cascade algorithm is the choice of the criteria
to switch from one optimizer to another. In the present case, a heuristic approach
was adopted based on numerical experiments. The criteria, as suggested in [186],
were: i) Switching from LM to EA: if, from one iteration to another, the relative
decrease in the quadratic objective function is less than 110
15
, or the maximum
admissible iteration number (predetermined value) is reached. ii) Switching from
EA to LM: if stagnation of more than 500 generations is observed or the relative
decrease in the quadratic objective function is less than 110
15
or the maximum
admissible iteration number (predetermined value) is reached. The obtained results
are summarized in Table 26 and the convergence evolution of the cascade
optimization strategies in the parameter identification are presented in Figure 97.
Table 26 Parameters achieved by the cascade optimization strategy
Parameter Value
 0.37
0.90
n 2.00
C
1
1.00
C
2
0.23
C
3
0.45
C
4
0.52
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 147
Figure 97 Convergence evolution of the cascade optimization strategies in the parameter
identification
Figure 98 shows the plot of the Relative Global Error, calculated for each direction
(using Equation (53)), for two situations, in order to compare the reference
simulation with the simplified model results, with and without considering the
interaction function, represented in the figure with filled and unfilled marks
respectively). By comparing the Relative Global Errors for the two situations the
error reduction in each direction is clear when the interaction function is
considered.
Figure 98 Relative Global Error of the simplified model results, with (filled marks) and
without (unfilled marks) interaction function, compared with the refined numerical results
0 500 1000 1500 2000
1.0x10
7
1.2x10
7
1.4x10
7
1.6x10
7
1.8x10
7
2.0x10
7
2.2x10
7
2.4x10
7
2.6x10
7
Q
u
a
d
r
a
t
i
c
e
r
r
o
r
Iterations/Generations
LM+EA+LM
0 20 40 60 80 100 120 140 160 180 200
1.0x10
7
1.2x10
7
1.4x10
7
1.6x10
7
1.8x10
7
2.0x10
7
2.2x10
7
2.4x10
7
2.6x10
7
LM
0 50 100 150 200 250
0
50
100
150
200
250
300
350
400
450
500
With interaction function optimized  o = 11.25
With interaction function optimized  o = 22.50
With interaction function optimized  o = 45.00
With interaction function optimized  o = 67.50
With interaction function optimized  o = 78.75
Without interaction function  o = 11.25
Without interaction function  o = 22.50
Without interaction function  o = 45.00
Without interaction function  o = 67.50
Without interaction function  o = 78.75
R
e
l
a
t
i
v
e
G
l
o
b
a
l
E
r
r
o
r
i
n
d
i
r
e
c
t
i
o
n
Y
(
%
)
Relative Global Error in direction X (%)
148 Chapter 6
Figure 99 includes a selected group of examples of pushover curves for different
columns and different pushover loading angles. In each plot, for both directions, the
obtained pushover curves are represented: i) by the refined numerical fibre model
(reference curves blue lines with square marks); ii) by the simplified model
without the biaxial bending interaction function (red lines with trianglular marks);
and iii) by the simplified model but with the interaction function (the optimised
solution green lines with diamond marks). Also, the examples in Figure 99
confirm the error reductions, in both column directions, obtained by adopting the
interaction function combined with the optimized scaling factor.
Figure 99 Examples of pushover curves for different columns and different pushover
loading angles for the refined numerical model, the simplified model without the biaxial
bending interaction function and the simplified model with the interaction function
0.00 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
300
350
400
450
500
550
M26
30x60cm
2
u = 0.2
As = 2.0% Ac
Pushover  11.25
o
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
0.00 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
300
350
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
M16
30x50cm
2
u = 0.1
As = 1% Ac
Pushover  22.5
o
0.00 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
300
350
400
M22
30x60cm
2
u = 0.1
As = 1.5% Ac
Pushover  45
o
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
0.00 0.01 0.02 0.03 0.04
0
50
100
150
200
250
300
350
M14
30x50cm
2
u = 0.2
As = 1.5% Ac
Pushover  45
o
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
0.00 0.01 0.02 0.03 0.04 0.05
0
25
50
75
100
125
150
M06
30x30cm
2
u = 0.3
As = 1.5% Ac
Pushover  67.5
o
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
0.00 0.01 0.02 0.03 0.04 0.05
0
50
100
150
200
250
300
350
400
M11
30x50cm
2
u = 0.2
As = 1% Ac
Pushover  11.25
o
B
a
s
e
S
h
e
a
r
(
k
N
)
Horizontal Displacement (m)
Fibre model (X direction)
Fibre model (Y direction)
Original (X)
Original (Y)
Optimized solution (X)
Optimized solution (Y)
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 149
6.5 Validation of the model with results from
cyclic tests
6.5.1 Introduction
For the validation of the proposed simplified model with interaction functions, the
experimental results of cyclic tests on twelve RC columns were used of the test
campaign presented in Chapters 3 and 4.
First, the uniaxial tests were modelled in order to obtain the primary skeleton
curves for each independent direction. Then, using the obtained curves, the biaxial
tests were simulated and the results of the numerical model with the interaction
function are compared with the experimental results.
6.5.2 Analysis of the results
For the validation of the proposed simplified interaction model, a trilinear envelope
curve was considered for the primary curve of each column direction, by bestfit
adjustment to the uniaxial experimental results.
For all uniaxially tested columns, the experimental results were well reproduced
with the adjusted uniaxial trilinear curves and with the hysteretic rules of the
original model, as represented in Figures 100, 103 and 106. In the figures the
following are plotted for each column: the numerical calculations with the
interaction model (blue); the experimental results (red); and the trilinear primary
curve (green).
As observed in these figures, and also before the strength degradation is difficult to
represent, particularly for the last cycles of the experimental response. However,
significant differences are only observed for demands corresponding to drifts greater
than 2.5%. The energy dissipation evolution is also well represented. Again, only
for the last cycles (associated with the differences in terms of strength
degradation), an overestimation of the dissipated energy is obtained with the
numerical model.
In order to validate the rules and parameters of the interaction model for the
simulation of the biaxial response of the tested RC columns, the trilinear curve
adjusted to the uniaxial test results was considered for the primary curve in each
direction.
The prediction of the experimental biaxial response obtained with the simplified
interaction model for the tested RC columns is represented in Figures 101, 102,
150 Chapter 6
104, 105, 107, 108 and 109. The same line legend is adopted as for the uniaxial
cases.
As can be observed, the maximum strength was properly obtained with the
simplified biaxial model. For the columns strong directions, a underestimation of
the maximum strength of 15% was observed, while overestimation of 25% is
reached in the weak direction. The unloading stiffness and the pinching effect were
reasonably reproduced in most cases.
The strength degradation was also reasonably approximately in the examples under
analysis. Only in the latter stages of the columns response, close to the columns
failure, considerable differences were detected in terms of strength degradation.
The evolution of the accumulated energy dissipation is reasonable well simulated
until noticeable strength degradation is observed. In the stronger column direction
an overestimation of around 20% was reached and in the weak direction the
underestimation is around 25%.
In general, a good agreement between the predicted numerical results and the
experimental hysteretic response was observed, indicating that the proposed
strategy may be suitable to simulate the response of columns to biaxial loading
based on uniaxial behaviour curves associated with properly calibrated coupling
interaction functions.
Figure 100 Baseshear versus drift of columns N05 and N06 Uniaxial tests
4 2 0 2 4
120
100
80
60
40
20
0
20
40
60
80
100
120
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
6 4 2 0 2 4 6
120
100
80
60
40
20
0
20
40
60
80
100
120
0
20
40
60
80
100
120
140
160
180
Ddrift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
PB01N05
X
Y
PB02N06
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 151
Figure 101 Baseshear versus drift of column N07 Biaxial test, rhombus displacement
pattern
Figure 102 Baseshear versus drift of column N04  Biaxial test, rhombus displacement
pattern
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
0
20
40
60
80
100
120
4 2 0 2 4
80
60
40
20
0
20
40
60
80
0
20
40
60
80
100
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
PB12N07
X
Y
PB12N07
X
Y
4.03.53.02.52.01.51.00.50.0 0.51.0 1.5 2.02.5 3.03.5 4.0
120
100
80
60
40
20
0
20
40
60
80
100
120
0
20
40
60
80
100
120
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
0
20
40
60
80
100
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
X
Y
PB12N08
PB12N08
X
Y
152 Chapter 6
Figure 103 Baseshear versus drift of columns N09 and N10 Uniaxial tests
Figure 104 Baseshear versus drift of column N11 Biaxial test, rhombus displacement
pattern
6 4 2 0 2 4 6
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
0
50
100
150
200
250
300
6 4 2 0 2 4 6
100
80
60
40
20
0
20
40
60
80
100
0
20
40
60
80
100
120
140
160
180
200
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
PB01N09
X
Y
X
Y
PB02N10
X
Y
X
Y
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
4 2 0 2 4
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
0
20
40
60
80
100
120
140
4 2 0 2 4
100
80
60
40
20
0
20
40
60
80
100
0
20
40
60
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
PB12N11
X
Y
X
Y
PB12N11
X
Y
X
Y
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 153
Figure 105 Baseshear versus drift of column N12 Biaxial test, quadrangular
displacement pattern
Figure 106 Baseshear versus drift of column N13 Uniaxial test
Figure 107 Baseshear versus drift of column N14 Biaxial test, rhombus displacement
pattern
4 2 0 2 4
180
150
120
90
60
30
0
30
60
90
120
150
180
0
20
40
60
80
100
120
140
4 2 0 2 4
100
80
60
40
20
0
20
40
60
80
100
0
20
40
60
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
PB12N12
X
Y
X
Y
PB12N12
X
Y
X
Y
PB12N12
X
Y
X
Y
6 4 2 0 2 4 6
60
40
20
0
20
40
60
0
20
40
60
80
100
120
140
160
180
200
Drift X (%)
S
h
e
a
r
X
(
k
N
)
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
PB01N13
X
Y
6 4 2 0 2 4 6
100
80
60
40
20
0
20
40
60
80
100
0
20
40
60
80
100
120
140
6 4 2 0 2 4 6
80
60
40
20
0
20
40
60
80
0
20
40
60
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
Drift X (%)
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
DriftY (%)
PB12N14
X
Y
PB12N14
X
Y
PB12N14
X
Y
154 Chapter 6
Figure 108 Baseshear versus drift of column N15 Biaxial test, quadrangular
displacement pattern
Figure 109 Baseshear versus drift of column N16 Biaxial test, circular displacement
pattern
6.6 Final comments
There is still a number of unsolved problems associated with modelling of RC
elements under biaxial loading. Simplified biaxial models may be adopted if they
can adequately reproduce the main characteristics of the elements response (such
as the strength and stiffness degradation, ductility, and energy dissipation
capacity) relative to the columns uniaxial response.
4 2 0 2 4
100
75
50
25
0
25
50
75
100
0
20
40
60
80
4 2 0 2 4
100
75
50
25
0
25
50
75
100
0
20
40
60
Drift X (%)
DriftY (%) Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
PB12N15
X
Y
PB12N15
X
Y
4 2 0 2 4
100
75
50
25
0
25
50
75
100
0
20
40
60
80
100
120
4 2 0 2 4
100
75
50
25
0
25
50
75
100
0
20
40
60
80
Drift Y (%)
Drift X (%)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
Step
A
c
c
u
m
u
l
a
t
i
v
e
h
y
s
t
e
r
e
s
i
s
d
i
s
s
i
p
a
t
e
d
e
n
e
r
g
y
(
k
N
.
m
)
S
h
e
a
r
Y
(
k
N
)
S
h
e
a
r
X
(
k
N
)
PB12N16
X
Y
X
PB12N16
X
Y
X
Simplified model for the nonlinear behaviour of RC columns under biaxial bending 155
In the present Chapter, a simplified interaction model for the response of RC
columns to biaxial loading is discribed, based on existing uniaxial models. The
proposed model corresponds to an upgrade of the existing CostaCosta uniaxial
hysteretic model, and adopts an interaction function based on the BoucWen
biaxial hysteretic model, coupling the two loading directions. The model
parameters were calibrated using optimization techniques, based on the results of a
parametric study on the tridimensional response of RC columns models with a
refined model. The validity of the proposed model was demonstrated through the
analytical simulation of biaxial tests on RC columns. The obtained numerical
results were adequate and proved the efficiency of the model, which is a simple tool
capable of reproducing the response of RC elements considering the biaxial
interaction.
The proposed simplified model can be a useful tool in the design and assessment of
RC structures where the response is dependent on biaxial bending of the elements.
This nonlinear model accounts for mechanical features such as hysteretic
behaviour rules, strength and stiffness degradation, and the pinching effect.
However, additional research is still necessary to objectively define the interaction
function parameters, which establish the coupling of the response in the two
loading directions. Moreover, the model application to experimental results
obtained by other authors should be made. The implementation of the proposed
model in a structural analysis program to obtain the response of RC multistorey
buildings (which strongly depends on the biaxial response of the columns) is
another important task that should be archived.
Chapter 7
Conclusions and Future
Research
This thesis aimed to study the response of RC columns for horizontal biaxial loads
combined with constant axial load. The work was focused in three main
directions: i) testing of 24 fullscale RC columns subjected to uniaxial or biaxial
cyclic horizontal demands with constant axial load; ii) comparative assessment of
the capacity of current nonlinear modelling strategies in the representation of the
response of RC columns; and iii) proposal of a simplified model to represent the
nonlinear behaviour of RC columns under monotonic and cyclic biaxial lateral
loads.
In the present chapter are summarised the main conclusions drawn from the work
developed within the present thesis. Suggestions for possible future research are
also made, intending to improve the knowledge concerning the influence of biaxial
demands on the behaviour of RC columns and its effect on the buildings response
when subjected to earthquake actions.
158 Chapter 7
7.1 Summary of conclusions
Based on the research conducted in this thesis, the following main conclusions can
be drawn:
 As discussed in Chapter 2, recent earthquakes proved that in many cases
the collapse, or major damage, observed in RC modern buildings are
associated to the failure of the columns.
 A review of the previous research work concerning the biaxial behaviour of
RC elements, presented in Chapter 2, underlines the limited number of
available test results, which justifies the need for additional experimental
research in this field.
 Chapters 3 and 4 focused on the presentation and analysis of the results of
the developed experimental campaign, consisting on the cyclic testing of 24
RC columns under uniaxial and biaxial loading. The results showed that, in
general, RC columns under biaxial earthquake loading achieve higher levels
of damage than for the case of uniaxial loading. For example, columns
biaxially loaded may reach the concrete cracking, concrete spalling and bar
buckling for drift levels between 50 and 75% of the corresponding value for
uniaxial demands. In the biaxial tests it was also verified that the
conventional rupture is anticipated relatively to the uniaxially loaded
columns, due to the pronounced strength degradation observed after
buckling of the longitudinal bars.
 From the experimental results, in terms of hysteretic sheardrift curves, the
influence of the biaxial loading was analysed, from which it was concluded
that: i) the reduction in the maximum strength due to biaxial loading was
approximately 8% in the columns strong direction and 20% in the weak
direction; ii) the ultimate ductility in columns subjected to biaxial loading
paths is approximately 50 to 75% lower in the column weak direction and
approximately 35% lower in the strong direction than for the uniaxial
loading; iii) for biaxial loadings, the strength degradation starts after a
demand corresponding to a displacement ductility factor of
approximately 3, and it was more pronounced than the strength
degradation observed in the corresponding uniaxial tests; iv) for the same
lateral deformation level, the biaxial loading tends to introduce larger
energy dissipation (for circular, rhombus and cruciform load paths) than
does the uniaxial loading.
 From the results on the tested biaxially columns, it was estimated the
equivalent viscous damping for each independent horizontal direction. It
was verified that the viscous damping highly depends on the biaxial load
Conclusions and Future Research 159
path. Two simplified expressions were proposed, based on the experimental
results, for the estimation of the equivalent biaxial damping in RC columns
subjected to biaxial loading, as a function of the ductility demand. It was
observed that the proposed expressions tend to overestimate the equivalent
damping, which justify further future research. However, they are
recognized as a first proposal of expressions for the estimation of the
equivalent damping of columns under biaxial loading.
 Based on the model proposed by Park and Ang for the Damage Index, two
expressions were proposed for the damage index of RC elements under
biaxial loading. These expressions were validated with the experimental
results. It is underlined that the proposed expressions should be calibrated
with results of tests of tridimensional buildings subjected to bidirectional
horizontal loadings.
 With the three refined numerical strategies studied (the distributed
inelasticity with force and displacement formulations and the lumped
plasticity), similar level of accuracy was verified in the representation of the
tested columns response. The three modelling strategies revealed limitations
in the representation of the strength degradation, particularly for larger
drift demands, which justifies the poor estimation of the energy dissipation
evolution with the numerical models.
 A simplified interaction model was proposed, based on the CostaCosta
uniaxial hysteretic model and the interaction function developed from the
BoucWen model. The validity of the proposed simplified model was
demonstrated through the analytical simulation of the response of the RC
columns biaxially tested. The obtained numerical results showed the model
efficiency, which encourages further validation to become a simple
tridimensional tool to simulate biaxially loaded RC elements.
160 Chapter 7
7.2 Future developments
As stated before, the number of bidirectional tests performed on RC columns
available in the literature is still limited. As a consequence, the influence of each
parameter affecting the columns response is not yet comprehensively understood.
The current research attempted to answer some of the open questions regarding the
influence of biaxial loading on the hysteretic behaviour of RC columns. However,
further investigation is considered necessary in order to achieve more
comprehensive conclusions and knowledge.
As experimental work, it is considered that the subjects requiring further
investigation are:
 To study the influence of other horizontal biaxial loading paths, and study
the response of columns for random biaxial hysteretic laws, obtained in
bidirectional pseudodynamic and shakingtable tests.
 To study the response of columns with different crosssections and
reinforcement steel arrangement, namely square columns with asymmetric
distribution of reinforcement steel; and a comparative study of sections with
the same percentage of reinforcement steel but with different distributions
of the bars in the section.
 To perform tests on columns with higher levels of axial load ratios, and to
analyse its influence in the damage evolution, ultimate point and failure
mechanism.
 To study the influence of the axial load variation simultaneously with the
horizontal displacement, representing more realistically the response of
building columns during earthquakes.
 To test columns with improved detailing in the plastic hinge region, as
recommend in actual seismic design codes, concluding about its influence in
the attenuation of the reinforcing steel bar buckling and in the increasing of
the ultimate ductility.
Regarding the nonlinear modelling, some aspects deserve further research,
particularly:
 Regarding the refined modelling, material models should be improved in
terms of the steel reinforcing bars buckling phenomena for large drift
demands, an important aspect to represent the strength degradation. The
proposals available in the literature depend on a series of parameters for
Conclusions and Future Research 161
which a precise calibration based in experimental results should be
developed.
 The equivalent viscous damping expression proposed should be validated
against results of nonlinear time history analyses of RC buildings.
 Regarding the simplified biaxial model proposed, further improvements and
developments are considered necessary, namely: i) simulation of
experimental results obtained by other authors; ii) development of a
sensitivity analysis of the simplified model to each parameter involved,
based on results obtained with refined models; iii) implementation of the
proposed simplified model in a structural analysis program to test the
applicability in the analysis of RC multistorey irregular buildings.
REFERENCES
[1] P. V. Patel, "Lessons from the recent earthquake" The Indian Concrete
Journal, pp. 600601, September 2001 2001.
[2] V. Bertero, "Lessons Learned from Recent Earthquakes and Research and
Implications for EarthquakeResistente Design of Building Structures in the
United States" Earthquake Spectra, vol. 2, p. 34, 1986.
[3] F. T. B. Zhaoa, T. Rossetto, "Field investigation on the performance of
building structures during the 12 May 2008 Wenchuan earthquake in
China" Engineering Structures, 2009.
[4] D. K. Miller, "Lessons learned from the Northridge earthquake" Engineering
Structures, vol. 20, p. 12, 1998.
[5] H. Sezen and K. J. E. A.S. Whittaker, K.M. Mosalam, "Performance of
reinforced concrete buildings during the August 17, 1999 Kocaeli, Turkey
earthquake, and seismic design and construction practise in Turkey"
Engineering Structures, vol. 25, p. 12, 2003.
[6] A. Dogangun, "Performance of reinforced concrete buildings during the May
1, 2003 Bingo l Earthquake in Turkey" Engineering Structures, vol. 26, p.
16, 2004.
[7] O. A. K. Adalier, "Structural engineering aspects of the June 27, 1998
AdanaCeyhan (Turkey) earthquake" Engineering Structures, vol. 23, p.
13, 2001.
[8] K.S. Pribadi and D.K.Rildova, "Learning from recent Indonesian
earthquakes: An overview to improve structural performance" presented at
the 14
th
World Conference on Earthquake Engineering, Beijing, China,
2008.
[9] A. R. Manafpour, "Bam earthquake, Iran: Lessons on the seismic behaviour
of building structures" presented at the 14
th
World Conference on
Earthquake Engineering, Beijing, China, 2008.
[10] N. VilaPouca, "Simulao numrica da resposta ssmica de elementos
laminares em beto armado" PhD, Faculdade de Engenharia, Universidade
do Porto, Porto, 2001.
164 References
[11] NCEER, "Preliminary Reports from the HyogoKen Nambu Earthquake of
January 17, 1995" State University of New York at Buffalo, Buffalo1995.
[12] R. Spence, "Saving lives in earthquakes: successes and failures in seismic
protection since 1960" Bulletin of Earthquake Engineering and Structural
Dynamics, vol. 5, p. 12, 2007.
[13] R. Bento, "Assessment of the Seismic Behaviour of Reinforced Concrete
Framed Structures: An Approach Based on Damage Indices" PhD, IST,
Lisbon, 1996.
[14] RSCCS, "Regulamento de Segurana das Construes Contra os Sismos,
DecretoLei n. 41658 de 31 de Maio de 1958", ed, 1958.
[15] RSEP, "Regulamento de Solicitaes em Edifcio e Pontes, DecretoLei n.
44041 de 18 de Novembro de 1961" 1961.
[16] RSA, "Regulamento de Segurana e Aces para Estruturas de Edifcios e
Pontes, DecretoLei n. 235/83, de 31 de Maio", ed, 1983.
[17] Eurocode 8: Design of structures for earthquake resistance  Part 13:
Strengthening and repair of buildings  European prEN 199813, B.
European Committee for Standardization, Belgium, 2003.
[18] F. Taucer, E. Spaconeand F. Filippou, "A fiber beamcolumn element for
seismic response analysis of reinforce concrete structures" University of
California, Berkeley UCB/EERC91/17, 1991.
[19] S. N. Bousias, T. B. Panagiotakosand M. N. Fardis, "Modelling of RC
members under cyclic biaxial flexure and axial force" Journal of Earthquake
Engineering, vol. 6, No 2, pp. 213238, 2002.
[20] X. Romo, A. Costaand R. Delgado, "New model for the inelastic biaxial
bending of reinforced concrete columns" in 13th World Conference on
Earthquake Engineering, Vancouver, B.C., Canada, 2004.
[21] M. Saatcioglu and G. Ozcebe, "Response of reinforced concrete columns to
simulated seismic loading" ACI Structural Journal, vol. no. 86S1, 1989.
[22] F. Qiu, W. Li, P. Panand J. Qian, "Experimental tests on RC columns
under biaxial quasistatic loading" Engrg Structures, vol. Vol. 24, pp. pp.
419428, 2002.
[23] S. Otani and "Rose School Shot Course on "Nonlinear analysis of reinforced
concrete buildings"", ed, 2002.
References 165
[24] H. Varum, "Seismic assessment, strengthening and repair of existing
buildings" PhD Thesis, Civil Engineering Department, University of Aveiro,
Aveiro, 2003.
[25] EASY. (1997, 27082010). Earthquake engineering slide information
system. Available: http://www.ikpir.com/easy/html/ang/INDEX.HTM
[26] H. Rodrigues, X. Romo, A. G. Costa, A. Arde, H. Varum, J. Guedes, R.
Vicente, A. A. Costaand E. Pauprio, "Sismo de L'Aquila de 6 de Abril de
2009. Ensinamentos para Portugal" presented at the 8 Congresso Nacional
de Sismologia e Engenharia Ssmica  SISMICA 2010, Aveiro, 2010.
[27] P. Ricci, F. De Lucaand G. Verderame, "6th April 2009 LAquila
earthquake, Italy: reinforced concrete building performance" Bulletin of
Earthquake Engineering, vol. 9, pp. 285305, 2011.
[28] H. Rodrigues, H. Varumand A. Costa, "Simplified MacroModel for Infill
Masonry Panels " Journal of Earthquake Engineering, Volume vol. 14, pp.
390  416 2010.
[29] R. Vicente, H. Rodrigues, A. Costa, H. Varumand J. A. R. M. d. Silva,
"Masonry Enclosure Walls: Lessons learnt from the recent Abruzzo
Earthquake" presented at the Resultados da pesquisa14th European
Conference On Earthquake Engineering, Ohrid, Macedonia, 2010.
[30] R. Vicente, H. Rodrigues, H. Varum, A. Costaand J. M. Silva,
"Performance of masonry enclosure walls: lessons learned from recent
earthquakes" Earthquake Engineering and Engineering Vibration, vol. 11,
2012.
[31] M. N. Fardis, "Seismic design issues for masonryinfilled RC frames"
presented at the Proceedings of the first European conference on
earthquake engineering and seismology, 2006.
[32] CEB, "RC frames under earthquake loading", Lausanne Bulletin 220, 1996.
[33] E. Coelho, "Comportamento ssmico de estruturas em prtico de beto
armado  Avaliao da resposta nolinear histertica" PhD Thesis, IST,
Lisbon, 1992.
[34] C. T. Vaz, "Comportamento ssmico de pontes com pilares de beto armado
 Verificao de segurana" PhD Thesis, Faculdade de Engenharia da
Universidade do Porto, Porto, 1992.
166 References
[35] L. Mendes and E. Coelho, "Seismic behaviour of RC structures Review of
the principal modelling strategies" presented at the First European
Conference on Earthquake Engineering and Seismology, Geneva,
Switzerland, 2006.
[36] R. Park and T. Paulay, Reinforced concrete structures: John Wiley & Sons,
1975.
[37] H. Rodrigues, A. Arde, H. Varumand A. Costa, "Behaviour of RC building
columns under cyclic loading: experimental study" Journal of Earthquake
and Tsunami, JET, World Scientific Publishing (in press). 2012.
[38] A. Arde, "Seismic assessment of reinforced concrete frame structures with
a new flexibility based element" PhD Thesis, Faculdade de Engenharia,
Universidade do Porto, Porto, 1997.
[39] S. Otani, V. W. T. Cheungand S. S. Lai, "Reinforced concrete columns
subjected to biaxial lateral load reversals" presented at the 7th World Conf.
on Earthq. Engrg V. 6, 525532., 1980.
[40] S. Otani, "Hysteresis models of reinforced concrete for earthquake response
analysis" Journal of Faculty of Engineering, vol. 36, No. 2, pp. 407 441,
1981.
[41] D. A. Ross and J. R. Yen, "Interactive design of reinforced concrete
columns with biaxial bending" ACI Journal, vol. 83, pp. 988993, 1986.
[42] J. L. Bonet, M. H. F. M. Barrosand M. L. Romero, "Comparative study of
analytical and numerical algorithms for designing reinforced concrete
sections under biaxial bending" Computers and Structures, vol. 84, pp.
21842193, 2006.
[43] S. N. Bousias, G. Verzelleti, M. N. Fardisand G. Magonette, "RC columns
in cyclic biaxial bending and axial load" presented at the 10th World Conf.
on Earthq. Engrg, 30413046, Madrid, 1992.
[44] K. Kawashima, H. Ogimoto, R. Hayakawaand G. Watanabe, "Effect of
bilateral excitation on the seismic performance of reinforced concrete bridge
columns" presented at the 8
th
U.S. National Confe. on Earthq. Engrg, paper
567, 2006.
[45] B. A. Lejano, "Investigation of biaxial bending of reinforced concrete
columns through fiber method modeling" Journal of Research in Science,
Computing, and Engineering, vol. 4, pp. 6173, 2007.
References 167
[46] K. N. Li, H. Aoyamaand S. Otani, "Reinforced concrete columns under
varying axial load and bidirectional horizontal load reversals" Proceedings
of the 9th World Conference on Earthquake Engineering, 1998.
[47] S. Low and J. P. Moehle, "Experimental study of reinforced concrete
columns subject to multiaxial cyclic loading" Earthq. Engrg. Res. Center,
vol. Rep. No UCB/EERC 8714, Univ. of California, Berkeley, Ca., 1987.
[48] S. N. Bousias, G. Verzeletti, M. N. Fardisand E. Gutierrez, "Loadpath
effects in column biaxial bending with axial force" Journal of Engineering
Mechanics, pp. 596605, 1995.
[49] H. Sezen, "Shear deformation model for reinforced concrete columns"
Structural Engineering and Mechanics, vol. 28, No. 1, pp. 3952, 2008.
[50] V. Bertero and E. Popov, "Seismic behaviour of ductile momentresisting
reinforced concrete frames" ACI Publication, vol. SP53, pp. 247292, 1977.
[51] M. E. Marante and J. FlrezLpez, "Model of damage for RC elements
subjected to biaxial bending" Engineering Structures, vol. 24, pp. 1141
1152, 2002.
[52] T. Paulay and M. J. N. Priestley, Seismic design of RC and masonry
buildings  John Wiley  ISBN 0471549150, 1992.
[53] S.Y. Chang, "Experimental Studies of Reinforced Concrete Bridge Columns
under Axial Load Plus Biaxial Bending" Journal of Structural Engineering,
pp. 1225, 2010.
[54] N. Suzuki, S. Otaniand Y. Kobayashi, "Threedimensional beamcolumn
subassemblages under bidirectional earthquake loading" presented at the
Eighth World Conference on Earthquake Engineering, San Francisco, 1984.
[55] G. Monti and C. Nuti, "Nonlinear cyclic behavior of reinforcing bars
including buckling" ASCE Journal of Structural Engineering, vol. 118, pp.
32683284, 1992.
[56] L. Li, J. B. Manderand R. P. Dhakal, "Bidirectional cyclic loading
experiment on a 3D beamcolumn joint designed for damage avoidance"
ASCE Journal of Structural Engineering, vol. 134, pp. 17331742, 2008.
[57] U. Akguzel and S. Pampanin, "Effects of variation of axial load and
bidirectional loading on seismic performance pf GFRP retrofitted reinforced
concrete exterior beamcolumn joints" Journal of Composites for
Construction, ASCE, vol. 14, pp. 94104, 2010.
168 References
[58] K. Takiguchi, S. Kokusho, K. Kobayashiand M. Kimura, "Response of RC
column to horizontal bidirectional deflection history" presented at the
Proc. 7th World conference on Earthquake engng, Istanbul, 1980.
[59] M. Berry, M. Parrishand M. Eberhard, "PEER Structural Performance
Database  Users Manual " Pacific Earthquake Engineering Research
Center, University of California, Berkeley2004.
[60] H. Takizawa, M. Aoyamaand . "Biaxial effects in modelling earthquake
response of RC structures" Earthquake Engineering and Structural
Dynamics, vol. V. 4, pp. 523552, 1976.
[61] R. Park, F. A. Zahnand T. J. Falconer, "Strength and ductility of
reinforced and prestressed concrete columns and piles under seismic
loading" presented at the World Conference on Earthquake Engineering,
San Francisco, CA., USA, 1984.
[62] K.N. Li, H. Aoyanaand S. Otani, "Reinforced concrete columns under
varying axial load and bidirectional lateral load reversals" presented at the
Proceedings of the Ninth World Conference on Earthquake Engineering,
TokyoKyoto, Japan, 1988.
[63] M. Saatcioglu and G. Ozcebe, "Response of reinforced concrete columns to
simulated seismic loading" ACI Structural Journal, vol. 86, pp. 312, 1989.
[64] F. A. Zahn, R. Parkand M. J. N. Priestly, "Strength and ductility of square
reinforced concrete column sections subjected to biaxial bending" ACI
Structural Journal, vol. 86, pp. 123131, 1989.
[65] J.K. Kim and S.S. Lee, "The behaviour of reinforced concrete columns
subjected to axial force and biaxial bending" Engineering Structures, vol.
23, pp. 15181528, 2000.
[66] K. Tsuno and R. Park, "Experimental study of reinforced concrete bridge
piers subjected to bidirectional quasistatic loading" Struct. Engrg
Structures, JSCE, vol. 21, No 1 11s26s, 2004.
[67] H. Bechtoula, S. Konoand F. Watanabe, "Experimental and analytical
investigations of seismic performance of cantilever reinforced concrete
columns under varying transverse and axial loads" Journal of Asian
Architecture and Building Engineering, vol. 4, pp. 467474, 2005.
[68] S.Y. Chang, "Experimental studies of reinforced concrete bridge columns
under axial load plus biaxial bending" Journal of Structural Engineering,
ASCE, pp. 1225, January 2010.
References 169
[69] T. Ohno and T. Nishioka, "An experimental study on energy absorption
capacity of columns in reinforced concrete structures" Proceedings of the
JSCE, Structural Engineering/Earthquake Engineering, vol. 1, No 2, pp.
137147, 1984.
[70] M. G. Oliva and R. W. Clough, "Biaxial seismic response of R/C frames"
Journal of Structural Engineering, vol. 113, pp. 12641281, 1987.
[71] E. Mola and P. Negro, "Fullscale PSD testing of the torsionally unbalanced
spear structure in the asbuilt and retrofiitted configurations" presented at
the Proc. of SPEAR International Workshop, 2005.
[72] J. Molina, P. Buchet, G. E. Magonette, O. Hubertand P. Negro,
"Bidirectional pseudodynamic technique for testing a threestorey reinforced
concrete building" presented at the 13
th
World Conference on Earthquake
Engineering Vancouver, B.C., Canada, 2004.
[73] R. T. Duarte, "The use of analytical methods in structural design for
earthquake resistance" in Experimental and Numerical Methods in
Earthquake Engineering, J. Dona and P. M. JonesEds.: Commission of the
European Communities. Joint Research Centre. Ispra Establishment, p.
149174, 1991.
[74] M. N. Fardis, "Membertype models for the nonlinear seismic response of
reinforced concrete structures" in Experimental and Numerical Methods in
Earthquake Engineering, D. a. P. M. Jones: Kluwer Academic Publishers,
Dordrecht, The Netherlands, 1991.
[75] P. Ceresa, L. Petriniand R. Pinho, "Flexureshear fiber beamcolumn
elements for modeling frame structures under seismic loading State of
the art" Journal of Earthquake Engineering, vol. 11, pp. 46  88, 2007.
[76] M. H. Scott, G. L. Fenves, F. McKennaand F. C. Filippou, "Software
patterns for nonlinear beamcolumn models" Journal of Structural
Engineering, vol. 134, pp. 562571, 2008.
[77] M. N. Aydinoglu and G. Onem, "Evaluation of analysis procedures for
seismic assessment and retrofit design" in Earthquake Engineering in
Europe, Geotechnical, Geolical and Earthquake Engineering, A. A. M.
Gravski: Springer Science, 2010.
[78] W. Clough, K. Benuskaand E. Wilson, "Inelastic earthquake response of tall
buildings" presented at the Proceedings of the Third World Conference on
Earthquake Engineering, New Zealand, 1965.
170 References
[79] C. Coelho, "Anlise ssmica de estruturas tridimensionais de edifcios de
beto armado" MSc, Departamento de Engenharia civil, FEUP, Porto,
1997.
[80] K. M. Galal and A. Ghobarah, "Flexural and shear hysteretic behaviour of
RC columns with variable axial load" Engineering Structures, vol. 25(11),
pp. 13531367, 2003.
[81] M. G. Sfakianakis and M. N. Fardis, "Bounding surface model for cyclic
biaxial bending of RC sections" Journal of Engineering Mechanics, vol. 117,
pp. 27482769, 1991.
[82] A. DAmbrisi and F. C. Filippou, "Modeling of cyclic shear behavior in RC
members" Journal of Structural Engineering, pp. 11431149, 1999.
[83] H. Sezen, "Seismic behavior and modeling of reinforced concrete building
columns" PhD, Engineering  Civil and Environmental Engineering,
University of California, Berkeley, 2000.
[84] F. C. Filippou and G. L. Fenves, "Methods of analysis for earthquake
resistant structures" in Earthquake Engineering  From Engineering
Seismology to PerformanceBased Engineering, Y. Bozorgnia and V. V.
Bertero, Eds., ed. Cambridge, United Kingdom: Cambridge University
Press, 2004.
[85] S. Otani, "Inelastic analysis of RC frame structures" Journal of the
Structural Division, ASCE, pp. 14331449 1974.
[86] A. Calabrese, "Numerical issues in distributed inelasticity modelling of RC
frame elements for seismic analysis" MsC, Istituto Universitario di Studi
Superiori di Pavia, Universit degli Studi di Pavia, Pavia, 2008.
[87] E. Spacone, V. Ciampiand F. Filippou, "A beam element for seismic
damage analysis" vol. UCB/EERC92/07, ed. University of California,
Berkeley: Earthquake Engineering Research Center, 1992.
[88] ATC40, "Seismic evaluation and retrofit of existing concrete buildings,
Applied Technology Council, Redwood City (CA)", ed, 1996.
[89] BSSC, " NEHRP Recommended provisions for seismic regulation of new
buildings and other structures (FEMA 450)", ed: Federal Emergency
Management Agency, Washington, D.C., 2003.
References 171
[90] FEMA356, "Prestandard and commentary for the seismic rehabilitation of
buildings", ed: Federal Emergency Management Agency, Washington (DC),
2000.
[91] M. Petrangeli, P. E. Pintoand V. Ciampi, "Fiber element for cyclic bending
and shear of RC structures. I: Theory" Journal of Engineering Mechanics,
vol. 125, pp. 9941001, 1999.
[92] D. Pecknold, "Inelastic structural response to 2D ground motion" ASCE J.
Eng. Mech. Div., vol. 100, pp. 949963, 1974.
[93] G. H. Powell and P. F. Chen, "3D beamcolumn element with generalized
plastic hinges" ASCE J. Eng. Mech. Div., V. 112(7), 627641, 1986.
[94] K. ElMandooh Galal and A. Ghobarah, "Flexural and shear hysteretic
behaviour of reinforced concrete columns with variable axial load"
Engineering Structures, vol. 25, pp. 13531367, 2003.
[95] Y. K. Wen, "Method for random vibration of hysteretic systems" ASCE J.
Eng. Mech. Div., vol. 102(EM2), pp. 249263, 1976.
[96] S. K. Kunnath and A. M. Reinhorn, "Model for inelastic biaxial bending
interaction of RC beamcolumns" ACI Structural Journal, vol. 87 pp. 284
291, 1990.
[97] F. Casciati, "Stochastic dynamics of hysteretic media" Structural Safety, V.
6(24), 259269, 1989.
[98] C. H. Wang and Y. K. Wen, "Evaluation of preNorthridge lowrise steel
buildings. I: Modelling " J. Struct. Eng., ASCE, V. 126(10), 11601168.,
2000.
[99] M. G. Sfakianakis and M. N. Fardis, "RC Column model for inelastic
seismic response analysis in 3D" Journal of Engineering Mechanics, vol.
117, pp. 27702787, 1991.
[100] M. E. Marante and J. FlrezLpez, "Threedimensional analysis of
reinforced concrete frames based on lumped damage mechanics"
International Journal of Solids and Structures, vol. 40, pp. 51095123, 2003.
[101] F. Mazza and M. Mazza, "A numerical model for the nonlinear seismic
analysis of threedimensional RC frames" in The 14
th
World Conference on
Earthquake Engineering, Beijing, China, 2008.
172 References
[102] J. P. MirandaGuedes, "Seismic behaviour of reinforced concrete bridges.
Modelling, numerical analysis and experimental assessment" PhD Thesis,
Universidade do Porto, Porto, 1997.
[103] J. Zhao and S. Sritharan, "Modelling of strain penetration effects in fibre
based analysis of reinforced concrete structures" ACI Structural Journal,
vol. 104 pp. 133141, 2007.
[104] A. Monteiro, "Simulao numrica e validao experimental do
comportamento cclico de pilares ocos de pontes de beto armado " PhD
Thesis, Universidade do Porto, Porto, 2009.
[105] S. S. Lai, G. T. Willand S. Otani, "Model for inelastic biaxial bending of
concrete members" Journal of Structural Engineering, vol. 110, pp. 2563
2584, 1984.
[106] S. S. Lai and G. T. Will, "R/C space frames with column axial force and
biaxial bending moment interactions" Journal of Structural Engineering,
vol. 112, pp. 15531572, 1986.
[107] M. Saiidi, G. Ghusnand Y. Jiang, "Fivespring element for biaxially bent
R/C columns" Journal of Structural Engineering, vol. 115, pp. 398416,
1989.
[108] M. Saiidi, "Hysteresis models for reinforced concrete" Journal of the
Structural Division, ASCE, vol. 108, pp. 10771087, 1982.
[109] Q. Zhang, X. Guand Q. Huang, "Computer simulation of 3dimensional
seismic response for RC frame structures" presented at the Joint
International Conference on Computing and Decision Making in Civil and
Building Engineering, Montral, Canada, 2006.
[110] R. Bouc, "Forced vibration of mechanical system with hysteresis" presented
at the 4th Conference on Nonlinear Oscillation, Prague, 1967.
[111] Y. J. Park, Y. K. Wenand A. H.S. Ang, "Random vibration of hysteretic
systems under bidirectional ground motions" Earthquake Engineering and
Structural Dynamics, vol. 14, pp. 543557, 1986.
[112] R. Delgado, P. Delgado, N. VilaPouca, A. Arde, P. Rochaand A. Costa,
"Shear effects on hollow section piers under seismic actions: experimental
and numerical analysis" Bulletin of Earthquake Engineering, vol. 7, pp. 377
389, 2009.
References 173
[113] P. Delgado, V. Rodrigues, P. Rocha, M. Santos, A. Arde, N. Vila Pouca,
A. Costaand R. Delgado, "Experimental Tests on Seismic Retrofit of RC
Piers" in 8NCEE  Eighth U.S. National Conference On Earthquake
Engineering, San Francisco, California, 2006.
[114] P. Delgado, P. Rocha, V. Rodrigues, A. Arde, N. V. Pouca, A. Costa, R.
Delgadoand M. Santos, "Experimental tests on seismic retrofit of RC
columns" in 1
st
US  Portugal International Workshop. Grand challenges in
earthquake engineering. 250 years after the 1755 Lisbon Earthquake.,
Lamego, 2005, pp. 16.1 16.10.
[115] V. Rodrigues, "Reforo ssmico de pilares de beto armado  Anlise e
validao experimental" Msc Thesis, Departamento de Engenharia Civl,
Faculdade de Engenharia da Universidade do Porto, Porto, 2006.
[116] P. Rocha, P. Delgado, V. Rodrigues, M. Santos, A. Arde, N. Vila Pouca,
A. Costaand R. Delgado, "Seismic rehabilitation of reinforced concrete
columns" in Proceedings of the First European Conference on Earthquake
Engineering and Seismology (a joint event of the 13
th
European Conference
on Earthquake Engineering & 30
th
General Assembly of the European
Seismological Commission), Geneva, Switzerland, 2006.
[117] P. Delgado, "Avaliao da Segurana Ssmica de Pontes" PhD Thesis PhD
Thesis, Departamento de Engenharia Civil, Faculdade de Engenharia da
Universidade do Porto, 2009.
[118] P. Rocha, "Reforo e Reparao de Pilares de Edifcios de Beto Armado
em Zonas Ssmicas " PhD, Faculdade de Engenharia, Universidade do
Porto, Porto, 2011.
[119] Concrete  Specification, performance, production and conformity
(Portuguese Version), NPEN206, 2000.
[120] EC2, "Design of concrete structure, part 11: general rules and rules for
buildings" European standard EN 199211. European Committee for
Standardization (CEN), Brussels, 2004.
[121] Regulamento de Estruturas de Beto Armado e PrEsforado, 1983.
[122] Tensile Testing of Metallic Materials  Part 1: Method of Test at Ambient
Temperature, NP EN 100021, 2006.
[123] H. Rodrigues, A. Arde, H. Varumand A.G. Costa, "Experimental
evaluation of rectangular reinforced Concrete column behaviour under
174 References
biaxial cyclic Loading" Earthquake Engineering and Structural Dynamics
(in press), John Wiley & Sons, Ltd. 2012.
[124] H. Rodrigues, A. Arde, H. Varumand A. Costa, "Damage evolution in
reinforced concrete columns subjected to biaxial loading" Bulletin of
Earthquake Engineering (submitted, under Review), 2012.
[125] H. Rodrigues, H. Varum, A. Ardeand A. Costa, "A comparative analysis of
energy dissipation and equivalent viscous damping of RC columns subjected
to uniaxial and biaxial loading" Engineering Structures, vol. 35, pp. 149
164, 2012.
[126] R. P. Dhakal and R. C. Fenwick, "Detailing of Plastic Hinges in Seismic
Design of Concrete Structures" ACI Structural Journal, vol. 105, pp. 740
749, 2008.
[127] S. Bae and O. Bayrak, "Plastic Hinge Length of Reinforced Concrete
Columns" ACI Structural Journal, vol. 105, pp. 290300, MayJune 2008.
[128] Y. J. Park, A. H. S. Angand Y. K. Wen, "Damagelimiting aseismic design
of buildings" Earthquake Spectra, vol. 3, No. 1, 1987.
[129] A. Ghobarah, "Performancebased design in earthquake engineering: state of
development" Engineering Structures, vol. 23, pp. 878884, 2001.
[130] SEAOCVision2000, "Performance based seismic engineering of buildings,
vols. I and II: Conceptual framework", ed: Structural Engineers Association
of California, Sacramento (CA), 1995.
[131] FEMA273, "NEHRP guidelines for the seismic rehabilitation of buildings;
FEMA 274, Commentary", ed: Federal Emergency Management Agency,
Washington (DC), 1996.
[132] H. Mutsuyoshi, A. Machida, W. Tanzoand N. Mashiko, "Inelastic seismic
response of RC bridge pier using pseudodynamic test method" Transactions
of the Japan Concrete Institute, vol. 16, pp. 265272, 1994.
[133] T. Paulay, "An estimation of displacement limits for ductile systems"
Earthquake Engineering and Structural Dynamics, vol. 31, pp. 583599,
2002.
[134] M. J. N. Priestley, G. M. Calviand M. J. Kowalsky, Direct Displacement
Based Seismic Design of Structures: IUSS Press, Pavia, 2007.
References 175
[135] M. Aschheim, "Seismic design based on the yield displacement" Earthquake
Spectra, vol. 18, pp. 581600, 2002.
[136] R. Park, "Evaluation of ductility of structures and structural assemblages
from laboratory testing" Bulletin of the New Zealand National Society for
Earthquake Engineering, vol. 22, 1989.
[137] U. Akguzel and S. Pampanin, "Effects of variation of axial load and
Bidirectional load on Seismic Performance of GFRP Retrofitted Reinforced
Concrete Exterior BeamColumn Joints" Journal of Composites for
Construction  ASCE, vol. 94104, 2010.
[138] J. Hoshikuma, S. Unjoh, K. Kawashimaand K. Nagaya, "A Ductility
Evaluation of Flexural Reinforced Concrete Column Bases on Loading
History and Plastic Curvature" Journal of Structural Engineering, ASCE,
vol. 442, pp. 877888, 1998.
[139] M. B. Atalay and J. Penzien, "The Seismic Behaviour of Critical Regions of
Reinforced Concrete Components as Influenced by Moment, Shear and
Axial Force" Earthq. Engrg. Res. Center, vol. Rep.No. UCB/EERC 7519,
University of California, Berkeley, 1975.
[140] D. Abrams, "Influence of axial force variation on flexural behavior of
reinforced concrete columns" ACI Struct. J., vol. 84, pp. 246254, 1987.
[141] A. Elmenshawi and T. Brown, "Hysteretic energy and damping capacity of
flexural elements constructed with different concrete strengths" Engineering
Structures, vol. 32, pp. 297305, 2010.
[142] T. S. Eom and H. G. Park, "Evaluation of energy dissipation of slender
reinforced concrete members and its applications" Engineering Structures,
vol. 32, pp. 28842893, 2010.
[143] Y. J. Park and H. S. Ang, "Seismic Damage Model for Reinforced Concrete"
ASCE  J Struct Eng, vol. 111(4), pp. 722739, 1985.
[144] C. K. Nmai and D. Darwin, "Cyclic behavior of lightly reinforced concrete
beams. University of Kansas Centre for Research, SM report no. 12." 1984.
[145] D. Darwin and C. K. Nmai, "Energy Dissipation in RC Beams Under Cyclic
Load" Journal of Structural Engineering, vol. 112, pp. 18291846, 1986.
[146] ACI31808, " Building code requirements for structural concrete (ACI 318
08) and commentary (ACI 318R08)", ed: USA: American Concrete
Institute, 2008.
176 References
[147] M. J. N. Priestley, "Displacementbased seismic assessment of reinforced
concrete buildings" Journal of Earthquake Engineering  Imperial College of
London, vol. 1, pp. 157192, 1997.
[148] M. J. N. Priestley, "Myths and Fallacies in Earthquake Engineering,
Revisited The Mallet Milne Lecture" IUSS Press, Pavia, Italy. , 2003.
[149] M. J. N. Prieltley, M. J. Calviand M. J. Kowalsky, DisplacementBased
Sismic Design of Structures: IUSS Press, 2007.
[150] E. Rosenblueth and I. Herrera, "On a kind of hysteretic damping" Journal
of Engineering Mechanics Division ASCE, vol. 90, pp. 3748, 1964.
[151] P. Gulkan and M. A. Sozen, "Inelastic Responses of Reinforced Concrete
Structures to Earthquakes Motions" Procedings of the ACI, Vol 71, No 12.
, 1974.
[152] M. J. Kowalsky, "Displacement based designA methodology for seismic
design applied to RC bridge columns" MSc Thesis, University of California,
San Diego, California, 1994.
[153] C. Blandon, "Equivalent Viscous Damping Equations for Direct
Displacement Based Design" Master, Rose School, Pavia, 2004.
[154] B. Stojadinovic and C. R. Thewalt, "Energy balanced hysteresis models"
College of Engineering, University of California at Berkeley1996.
[155] Y. Lu, H. Hao, P. G. Carydisand H. Mouzakis, "Seismic performance of RC
frames~designed for three different ductility levels" Eng Struct vol. 25, pp.
537547, 2001.
[156] M. J. Kowalsky and H. Dwairi, "nvestigation of Jacobsen's Equivalent
Viscous Damping Approach As Applied To DisplacementBased Seismic
Design" presented at the Proceedings of the 13th World Conference on
Earthquake Engineering, Vancouver, Canada, 2004.
[157] M. J. N. Priesltley, M. J. Calviand M. J. Kowalsky, DisplacementBased
Sismic Design of Structures: IUSS Press, 2007.
[158] M. Fardis and T. Panagiotakos, "Hysteretic damping of reinforced concrete
elements" presented at the 11
th
World Conference on Earthquake Eng,
Acapulco, Mexico, 1996.
[159] Eurocode 8: Design of structures for earthquake resistance  Part 13:
Strengthening and repair of buildings  European prEN 199813, B.
European Committee for Standardization, Belgium, 2005.
References 177
[160] X. Romo, R. Delgado, J. Guedesand A. Costa, "A comparative application
of different EC83 procedures for the seismic safety assessment of existing
structures" Bull Earthquake Eng, vol. 8, pp. 91118, 2010.
[161] G. M. Verderame, P. Ricci, G. Manfrediand E. Cosenza, "Ultimate chord
rotation of RC columns with smooth bars: some considerations about EC8
prescriptions" Bull Earthquake Eng, 2010.
[162] B. Acun and H. Sucuoglu, "Performance of Reinforced Concrete Columns
Designed for Flexure under Severe Displacement Cycles" ACI Structural
Journal, vol. 103, pp. 364371, 2010.
[163] V. Popa, D. Cotofana, R. Pascuand A. Papurcu, "Study on the
Displacement Capacity of Bending Controlled Reinforced Concrete
Columns" presented at the 14
th
European Conference on Earthquake
Engineering, Ohrid, Macedonia, 2010.
[164] S. Pujol, M. A. Sozenand J. A. Ramirez, "Displacement history effects on
drift capacity of reinforced concrete columns" ACI Struct Journal, vol. 103,
pp. 253262, 2006.
[165] M. Rozman and P. Fajfar, "Seismic response of a RC frame building
designed according to old and modern practices" Bull Earthquake Eng, vol.
7, pp. 779799, 2009.
[166] M. M. Hachem, S. A. Mahinand J. P. Moehle, "Performance of Circular
Reinforced Concrete Bridge Columns under Bidirectional Earthquake
Loading", PEER 2003/06  UCB/ENG9374, University of California,
Berkeley, 2003.
[167] J. E. Stephens and J. P. T. Yao, "Damage assessment using response
measurements" ASCE Journal of Structural Engineering, vol. 113, pp. 787
801, 1987.
[168] M. S. Williams and R. G. Sexsmith, "Seismic Damage Indices for Concrete
Structures: A StateoftheArt Review" Earthquake Spectra, vol. 11, pp.
319349, 1995.
[169] Y. S. Chung, C. Meyerand M. Shinozuka, "Seismic damage assessment of
reinforced concrete members" NCEER870022, National Center for
Earthquake Engineering Research, Buffalo, N.Y., 1987.
[170] M. N. Fardis, S. N. Economu, A. N. Antoniou, P. J. Komodromosand M.
G. Sfakianakis, "Damage Measures and Failure Criteria  Part I"
Contribuition of University of Patras, Final Report of Cooperative Research
178 References
on the Seismic Response of Reinforced Concrete Structures  2
nd
Phase,
LNEC, Lisboa, 1993.
[171] S. K. Kunnath, A. M. Reinhornand Y. J. Park, "Analytical modelling of
inelastic seismic response of RC structures" ASCE, Journal of Structural
Engineering, vol. 116, N4, 1990.
[172] H. Varum, H. Rodriguesand A. Costa, "Numerical Models for Simulation of
the Seismic Behaviour of RC Structures  A Case Study" in Numerical
Modeling of Coupled Phenomena in Science and Engineering  Practical
Uses and Examples J. B. Mario Csar Surez Arriaga, Francisco Javier
DominguezMota Chapter 11 Taylor & Francis, 2008, pp. 141154.
[173] H. Rodrigues, H. Varum, A. Ardeand A. Costa, "Comparative efficiency
analysis of different nonlinear modelling strategies to simulate the biaxial
response of RC columns" Earthquake Engineering and Engineering
Vibration (Under Review), 2012.
[174] I.E. Bal, H. Craowleyand R. Pinho, "DisplacementBased Earthquake Loss
Assessment: Method Development and Application to Turkish Building
Stock", Rose School, IUSS Press, Research Reporto No. ROSE2010/02,
Pavia, Italy, 2010.
[175] A. Calabrese, J. P. Almeidaand R. Pinho, "Numerical Issues in Distributed
Inelasticity Modeling of RC Frame Elements for Seismic Analysis" Journal
of Earthquake Engineering, vol. 14, pp. 3868, 2010.
[176] M. J. N. Priestley and R. Park, "Strength and Ductility of Concrete Bridge
Columns Under Seismic Loading" ACI Structural Journal, vol. 84, pp. 61
76, 1987.
[177] H. Rodrigues, A. Arde, H. Varumand A. G. Costa, "Experimental study on
the biaxial bending cyclic behaviour of RC columns" presented at the 14th
European Conference on Earthquake Engineering Ohrid, Republic of
Macedonia, 2010.
[178] J.B. Mander, M.J.N. Priestleyand R. Park, "Theoretical stressstrain model
for confined concrete" Journal of Structural Engineering, vol. 114, pp. 1804
1826, 1988.
[179] A. C. Costa and A. G. Costa, "Hysteretic model of forcedisplacement
relationships for seismic analysis of structures" National Laboratory for
Civil Engineering, Lisbon, 1987.
References 179
[180] H. Rodrigues, X. Romo, A. AndradeCampos, H. Varum, A. Ardeand A.
G. Costa, "Simplified hysteretic model for the representation of the biaxial
bending response of RC columns" Engineering Structures (Under Review),
2012.
[181] T. Takeda, M. A. Sozenand N. N. Nielsen, "Reinforced Concrete Response
to Simulated Earthquakes" ASCE, Journal of the Structural Division, vol.
96, 1970.
[182] M. L. Wang and S. P. Shah, "Reinforced concrete hysteresis model based on
the damage concept" Earthquake Engineering & Structural Dynamics, vol.
15, pp. 9931003, 1987.
[183] R. Bouc, "Modle mathmatique dhystrsis", Acustica, vol. 24, pp. 1625,
1971.
[184] A. AndradeCampos, S. Thuillier, P. Pilvinand F. TeixeiraDias, "On the
determination of material parameters for internal variable thermoelastic
viscoplastic constitutive models" International Journal of Plasticity, vol. 23,
pp. 13491379, 2007.
[185] O. T. Bruhns and D. K. Anding, "On the simultaneous estimation of model
parameters used in constitutive laws for inelastic material behaviour"
International Journal of Plasticity, vol. 15, pp. 13111340, 1999.
[186] A. AndradeCampos, P. Pilvin, J. Simesand F. TeixeiraDias, "Software
Development for Inverse Determination of Constitutive Model Parameters"
in Software Engineering: New Research: Nova Science Publishers, Inc.,
2009.
[187] A. AndradeCampos, "Development of an Optimization Framework for
Parameter Identification and Shape Optimization Problems in Engineering"
International Journal of Manufacturing, Materials, and Mechanical
Engineering (IJMMME), vol. 1, pp. 5779, 2011.
[188] SeismoSoft, "SeismoStruc A computer program for static and dynamic
nonlinear analysis of framed structures [online]", ed: Available from URL:
http://www.seismosoft.com, 2004.
Appendix A
Experimental results
A.1 Introduction
In the present Appendix are presented the global results in terms of
hysteresis ShearDrift curves, biaxial displacement and force path
measured, and energy dissipation obtained for each column and presented
in Chapter 4.
Experimental results A3
A.2 Sheardrift curves
Figure A1  ShearDrift hysteretic curves for columns N01 to N04
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
Displacement Ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB01N01
10 8 6 4 2 0 2 4 6 8 10
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB02N02
10 8 6 4 2 0 2 4 6 8 10
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N04
10 8 6 4 2 0 2 4 6 8 10
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N04
X
Y
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
Displacement ductility factor
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N03
10 8 6 4 2 0 2 4 6 8 10
Displacement ductility factor
4 3 2 1 0 1 2 3 4
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N03
X
Y
10 8 6 4 2 0 2 4 6 8 10
Displacement ductility factor
A4 Appendix A
Figure A2  ShearDrift hysteretic curves for columns N05 to N08 and N17
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB01N05
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB02N06
20 16 12 8 4 0 4 8 12 16 20
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N07
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N07
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N08
18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18
5 4 3 2 1 0 1 2 3 4 5
120
100
80
60
40
20
0
20
40
60
80
100
120
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N08
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
X
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N07
10 8 6 4 2 0 2 4 6 8 10
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N17
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
Displacement ductility factor
Experimental results A5
Figure A3  ShearDrift hysteretic curves for columns N09 to N12 and N18
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB01N09
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB02N10
6 4 2 0 2 4 6
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N11
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N11
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N12
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N12
8 6 4 2 0 2 4 6 8
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
X
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N18
12 10 8 6 4 2 0 2 4 6 8 10 12
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
X
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N18
18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18
A6 Appendix A
Figure A4  ShearDrift hysteretic curves for columns N013 to N16
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB01N13
6 4 2 0 2 4 6
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N14
6 4 2 0 2 4 6
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N14
6 4 2 0 2 4 6
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N15
6 4 2 0 2 4 6
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N15
8 6 4 2 0 2 4 6 8
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
X
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N16
6 4 2 0 2 4 6
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N16
6 4 2 0 2 4 6
Experimental results A7
Figure A5  ShearDrift hysteretic curves for columns N19, N20, N23 and N24
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N19
6 4 2 0 2 4 6
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N19
8 6 4 2 0 2 4 6 8
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N20
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
5 4 3 2 1 0 1 2 3 4 5
200
150
100
50
0
50
100
150
200
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N20
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
3 2 1 0 1 2 3
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N23
12 10 8 6 4 2 0 2 4 6 8 10 12
3 2 1 0 1 2 3
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N23
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
3 2 1 0 1 2 3
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N24
10 8 6 4 2 0 2 4 6 8 10
3 2 1 0 1 2 3
100
75
50
25
0
25
50
75
100
Displacement ductility factor
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N24
10 8 6 4 2 0 2 4 6 8 10
A8 Appendix A
Figure A6  ShearDrift hysteretic curves for columns N21, N22
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N21
14 12 10 8 6 4 2 0 2 4 6 8 10 12 14
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N21
16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
X
(
k
N
)
Drift X (%)
PB12N22
18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18
Displacement ductility factor
5 4 3 2 1 0 1 2 3 4 5
200
160
120
80
40
0
40
80
120
160
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Drift Y (%)
PB12N22
20 15 10 5 0 5 10 15 20
Displacement ductility factor
Experimental results A9
A.3 Force and displacement paths
Figure A7 Measured displacement and force path for biaxial tests N03 and N04
60 40 20 0 20 40 60
60
40
20
0
20
40
60
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N03
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N03
70 35 0 35 70
70
35
0
35
70
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N04
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N04
X
Y
A10 Appendix A
Figure A8 Measured displacement and force path for biaxial tests N07, N08 and N17
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N07
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N07
X
Y
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N08
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N08
X
Y
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
X
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N17
140 120 100 80 60 40 20 0 20 40 60 80 100 120 140
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N17
X
Y
Experimental results A11
Figure A9 Measured displacement and force path for biaxial tests N11, N12 and N18
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N11
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
PB12N11
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N12
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N12
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N18
X
Y
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
X
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N12
A12 Appendix A
Figure A10 Measured displacement and force path for biaxial tests N14, N15 and N16
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N14
100 75 50 25 0 25 50 75 100
100
75
50
25
0
25
50
75
100
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
PB12N14
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N15
100 75 50 25 0 25 50 75 100
100
75
50
25
0
25
50
75
100
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
PB12N15
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
X
Y
PB12N16
100 75 50 25 0 25 50 75 100
100
75
50
25
0
25
50
75
100
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
X
Y
PB12N16
Experimental results A13
Figure A11 Measured displacement and force path for biaxial tests N14, N15 and N16
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N19
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
PB12N19
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
X
Y
PB12N20
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
PB12N20
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
PB12N23
X
Y
100 50 0 50 100
100
50
0
50
100
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N23
X
Y
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
PB12N24
X
Y
100 50 0 50 100
100
50
0
50
100
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N24
X
Y
A14 Appendix A
Figure A12 Measured displacement and force path for biaxial tests N21 and N22
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
PB12N21
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N21
X
Y
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
H
o
r
i
z
o
n
t
a
l
d
i
s
p
l
a
c
e
m
e
n
t

Y
d
i
r
e
c
t
i
o
n
(
m
m
)
Horizontal displacement  X direction (mm)
PB12N22
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
PB12N22
X
Y
Experimenta
A.4 I
al results
Individu
Figure A13
ual diss
3 Individua
sipated
al energy diss
energy
sipation for sp
y
pecimen N011 to N04
A15
A16
Figure A14 Individu ual energy disssipation for specimen N0 05 to N08 and
Appendix A
d N17
Experimenta
Fig
al results
gure A15 In ndividual eneergy dissipati ion for specim men N09 to NN12 and N18
A17
A18
Figurre A16 Ind dividual energgy dissipation n for specimen N13 to N1
Appendix A
6
Experimenta
al results
Figure A17 7 Individuaal energy diss sipation for sp pecimen N199 to N24
A19
Appendix B
Numerical simulations
B.1 Introduction
In the present Appendix are presented the global results in terms of
hysteresis ShearDrift curves, Energy dissipation, and biaxial force path
dissipation for each column obtained with each modelling strategy
presented in Chapter 5 and compared with the experimental results
obtained and presented in Chapter 4 and Appendix A.
Numerical simulations B3
B.2 Distributed inelasticity and displacement
based formulation
Figure B1  ShearDrift hysteretic curves for columns N01 to N04
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0
25
50
75
100
0
25
50
75
100
dissipated
energy
(kN.m)
D
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N02
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N03
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
PB01N04
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B4 Appendix B
Figure B2  ShearDrift hysteretic curves for columns N05 to N08 and N17
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N05
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
50
100
150
200
250
300
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N08
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N08
0
50
100
150
200
250
300
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N17
0
50
100
150
200
250
300
0
50
100
150
200
250
300
F
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N17
0
50
100
150
200
250
300
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B5
Figure B3  ShearDrift hysteretic curves for columns N09 to N12 and N18
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N09
0
50
100
150
200
250
300
350
400
450
500
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N10
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B6 Appendix B
Figure B4  ShearDrift hysteretic curves for columns N013 to N16
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N13
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N15
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N15
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N16
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N16
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
dissipated
energy
(kN.m)
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B7
Figure B5  ShearDrift hysteretic curves for columns N19, N20, N23 and N24
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B8 Appendix B
Figure B6  ShearDrift hysteretic curves for columns N21 and N22
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B9
Figure B7 Biaxial force path for biaxial tests N03 and N04
Figure B8 Biaxial force path for biaxial tests N07, N08 and N17
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
X
Y
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B10 Appendix B
Figure B9 Biaxial force path for biaxial tests N11, N12 and N18
Figure B10 Biaxial force path for biaxial tests N14, N15 and N16
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N11
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N12
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N18
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N14
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N15
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N16
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B11
Figure B11 Biaxial force path for biaxial tests N14, N15 and N16
Figure B12 Biaxial force path for biaxial tests N21 and N22
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N19
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N20
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N23
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N24
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N21
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N22
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B13
B.3 Distributed inelasticity and force based
formulation
Figure B13  ShearDrift hysteretic curves for columns N01 to N04
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0
25
50
75
100
0
25
50
75
100
dissipated
energy
(kN.m)
D
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N02
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N03
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
PB01N04
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B14 Appendix B
Figure B14  ShearDrift hysteretic curves for columns N05 to N08 and N17
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N05
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
50
100
150
200
250
300
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N08
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N08
0
50
100
150
200
250
300
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N17
0
50
100
150
200
250
300
0
50
100
150
200
250
300
F
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N17
0
50
100
150
200
250
300
dissipated
energy
(kN.m)
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B15
Figure B15  ShearDrift hysteretic curves for columns N09 to N12 and N18
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N09
0
50
100
150
200
250
300
350
400
450
500
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N10
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B16 Appendix B
Figure B16  ShearDrift hysteretic curves for columns N013 to N16
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N13
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N15
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N15
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N16
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N16
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
dissipated
energy
(kN.m)
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B17
Figure B17  ShearDrift hysteretic curves for columns N19, N20, N23 and N24
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B18 Appendix B
Figure B18  ShearDrift curves for columns N21 and N22
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B19
Figure B19 Biaxial force path for biaxial tests N03 and N04
Figure B20 Biaxial force path for biaxial tests N07, N08 and N17
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
X
Y
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B20 Appendix B
Figure B21 Biaxial force path for biaxial tests N11, N12 and N18
Figure B22 Biaxial force path for biaxial tests N14, N15 and N16
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N11
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N12
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N18
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N14
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N15
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N16
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B21
Figure B23 Biaxial force path for biaxial tests N14, N15 and N16
Figure B24 Biaxial force path for biaxial tests N21 and N22
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N19
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N20
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N23
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N24
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N21
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N22
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
23 Appendix B
B.4 Lumpedplasticity element
Figure B25  ShearDrift hysteretic curves for columns N01 to N04
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N01
0
25
50
75
100
0
25
50
75
100
dissipated
energy
(kN.m)
D
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N02
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N03
0
25
50
75
100
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
PB01N04
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
80
60
40
20
0
20
40
60
80
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N04
X
Y
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B24 Appendix B
Figure B26  ShearDrift hysteretic curves for columns N05 to N08 and N17
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N05
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N06
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N07
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N08
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
50
100
150
200
250
300
0
50
100
150
200
250
300
D
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
PB12N07
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N08
0
50
100
150
200
250
300
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N17
0
50
100
150
200
250
300
0
50
100
150
200
250
300
F
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
120
100
80
60
40
20
0
20
40
60
80
100
120
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N17
0
50
100
150
200
250
300
dissipated
energy
(kN.m)
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B25
Figure B27  ShearDrift hysteretic curves for columns N09 to N12 and N18
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB01N09
0
50
100
150
200
250
300
350
400
450
500
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB02N10
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N11
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N12
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
Y
PB12N12
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N18
0
50
100
150
200
250
300
350
400
450
500
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B26 Appendix B
Figure B28  ShearDrift hysteretic curves for columns N013 to N16
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement (m)
PB01N13
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N14
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N15
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N15
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N16
0
25
50
75
100
125
150
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
PB12N16
X
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
0
25
50
75
100
dissipated
energy
(kN.m)
Step
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B27
Figure B29  ShearDrift hysteretic curves for columns N19, N20, N23 and N24
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N19
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
200
150
100
50
0
50
100
150
200
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N20
0
25
50
75
100
125
150
175
200
dissipated
energy
(kN.m)
Step
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N23
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
100
80
60
40
20
0
20
40
60
80
100
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N24
0
25
50
75
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B28 Appendix B
Figure B30  ShearDrift hysteretic curves for columns N21 and N22
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N21
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
X
(
k
N
)
Displacement X (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
0.08 0.06 0.04 0.02 0.00 0.02 0.04 0.06 0.08
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
X
Y
S
h
e
a
r
Y
(
k
N
)
Displacement Y (m)
PB12N22
0
25
50
75
100
Step
dissipated
energy
(kN.m)
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B29
Figure B31 Biaxial force path for biaxial tests N03 and N04
Figure B32 Biaxial force path for biaxial tests N07, N08 and N17
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
X
Y
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
80 60 40 20 0 20 40 60 80
80
60
40
20
0
20
40
60
80
PB12N03
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
120 100 80 60 40 20 0 20 40 60 80 100 120
120
100
80
60
40
20
0
20
40
60
80
100
120
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
B30 Appendix B
Figure B33 Biaxial force path for biaxial tests N11, N12 and N18
Figure B34 Biaxial force path for biaxial tests N14, N15 and N16
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N11
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N12
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 200
200
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
200
PB12N18
X
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N14
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N15
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N16
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Numerical simulations B31
Figure B35 Biaxial force path for biaxial tests N14, N15 and N16
Figure B36 Biaxial force path for biaxial tests N21 and N22
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N19
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
200 150 100 50 0 50 100 150 200
200
150
100
50
0
50
100
150
200 PB12N20
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N23
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
100 80 60 40 20 0 20 40 60 80 100
100
80
60
40
20
0
20
40
60
80
100
PB12N24
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N21
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180
180
160
140
120
100
80
60
40
20
0
20
40
60
80
100
120
140
160
180
PB12N22
S
h
e
a
r
Y
(
k
N
)
Shear X (kN)
X
Y
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Experimental
Numeric
Appendix C
Publications in the framework
of the PhD Thesis
C.1 Introduction
This appendix list the papers (submitted for publication or published in
journals) and articles presented in conferences, developed in the
framework of the PhD thesis.
Publications in the framework of the PhD Thesis C3
C.2 Papers in international scientific journals
with referee
H. Rodrigues, A. Arde, H. Varum, and A. Costa, "Damage
evolution in reinforced concrete columns subjected to biaxial
loading" Bulletin of Earthquake Engineering (submitted in March
2012, under review).
H. Rodrigues, H. Varum, A. Arde, and A. Costa, "Comparative
efficiency analysis of different nonlinear modelling strategies to
simulate the biaxial response of RC columns" Earthquake
Engineering and Engineering Vibration (submitted in March 2012,
under review).
H. Rodrigues, X. Romo, A. AndradeCampos, H. Varum, A.
Arde, and A. G. Costa, "Simplified hysteretic model for the
representation of the biaxial bending response of RC columns"
Engineering Structures (submitted in March 2012, under review).
H. Rodrigues, A. Arde, H. Varum, and A. Costa, "Experimental
evaluation of rectangular reinforced concrete columns behaviour
under biaxial cyclic loading"  Earthquake Engineering & Structural
Dynamics, John Wiley & Sons, Ltd., ISSN 10969845, Manuscript
ID: EQE110165 (accepted for publication in April 2012, in press).
H. Rodrigues, H. Varum, Arde, A. and A. Costa  "A
comparative analysis of energy dissipation and equivalent viscous
damping of RC columns subjected to uniaxial and biaxial loading",
Engineering Structures, Volume 35, February 2012, Pages
149164, ISSN 01410296, 10.1016/j.engstruct.2011.11.014.
H. Rodrigues, A. Arde, H. Varum, and A. Costa  "Behaviour of
RC building columns under cyclic loading: Experimental study" 
Journal of Earthquake and Tsunami, JET, World Scientific
Publishing (accepted for publication in October 2010  to be
published in June 2012).
B4 Appendix C
C.3 Papers in international conferences
H. Rodrigues, A. Arde, H. Varum , A.G. Costa (2012) "Energy
dissipation and equivalent damping of RC columns subjected to
biaxial bending: an investigation based in experimental results"
15WCEE, 15
th
World Conference on Earthquake Engineering, to be
held in Lisbon, Portugal, September 2012 (abstract accepted).
H. Rodrigues, X. Romo, A. AndradeCampos, H. Varum, A.
Arde, and A. G. Costa, "Simplified model for the nonlinear
behaviour representation of reinforced concrete columns under
biaxial bending results" 15WCEE, 15
th
World Conference on
Earthquake Engineering, to be held in Lisbon, Portugal, September
2012 (abstract accepted).
H. Rodrigues, H. Varum, A. Arde, A.G. Costa  "Comparison of
different modeling strategies for the representation of non linear
response of RC columns subjected to biaxial loading"
International Conference on Recent Advances in Nonlinear Models
Structural Concrete Applications, CoRAN 2011, Edited by H.
Barros, R. Faria, C. Pina and C. Ferreira, pp. 339353, ISBN 978
9729652479, Coimbra, Portugal, 2425 November 2011.
H. Rodrigues, A. Arde, H. Varum, A.G. Costa "Experimental
study on the biaxial bending cyclic behaviour of RC columns"
ECEES  14
th
European Conference on Earthquake Engineering 
Ohrid, Republic of Macedonia, September 2010.
C.4 Papers in national conferences
H. Rodrigues, A. Arde, H. Varum, A.G. Costa "Comportamento
cclico de pilares de beto armado sujeitos Flexo BiAxial:
Estudo Experimental"  BE2010 Encontro Nacional Beto
Estrutural  LNEC, Lisbon, November, 2010.
H. Rodrigues, A. Arde, H. Varum, A.G. Costa "Anlise do
comportamento de pilares rectangulares de beto armado sujeitos a
solicitaes bidireccionais"  ISBN: 9789899569546  8 Congresso
Nacional de Sismologia e Engenharia Ssmica  SISMICA 2010 
Universidade de Aveiro, Aveiro, October, 2010.