You are on page 1of 188

Istituto Universitario Università degli

di Studi Superiori Studi di Pavia

EUROPEAN SCHOOL OF ADVANCED STUDIES IN REDUCTION OF SEISMIC RISK


ROSE SCHOOL

EVALUATION OF THE SEISMIC RESPONSE OF EXISTING R.C.

FRAME BUILDINGS WITH MASONRY INFILLS

A Dissertation Submitted in
Partial Fulfilment of the Requirements for the
Master Degree in Earthquake Engineering

By

Mario Galli

Supervisors:
Prof. GUIDO MAGENES
Prof. STEFANO PAMPANIN

December, 2006
The dissertation entitled “Evaluation of the seismic response of existing R.C. frame buildings
with masonry infills”, by Mario Galli, has been approved in partial fulfilment of the
requirements for the Master Degree in Earthquake Engineering.

Prof. Guido Magenes …… … ………__

Prof. Stefano Pampanin __………… … ……


Abstract

ABSTRACT

Comprehensive experimental-analytical studies on the seismic vulnerability of existing reinforced concrete


frame buildings, designed for gravity-loads only as typically found in most seismic prone countries before
the introduction of adequate seismic design code provisions, confirmed the inherent weaknesses of these
systems, due to inadequate detailing and a general lack of capacity design principles.
Controversial effects on the global inelastic mechanism can be expected depending on the infills properties
(mechanical characteristic and distribution) and the joint damage mechanism.
In this contribution, the interaction between un-reinforced masonry infills and r.c. frame systems, when
appropriately considering the joint zone non-linear behavior, is investigated through pushover and non-linear
time-history analyses on 2-D and 3-D multi-storey frame systems under uni-directional or bi-directional
input motions.
A simplified and reliable analytical model based on a concentrated plasticity approach and validated on
different experimental tests on beam-column joints and frame systems (with and without infills) is adopted
and proposed for extensive studies on seismic vulnerability of existing buildings.
The presence of infills can guarantee higher stiffness and strength, reducing the inter-storey drift demand,
while increasing the maximum floor accelerations. A further positive influence of the infills can be
recognized in the reduction of column interstorey shear contribution as well as in the possible delay of a soft-
storey mechanism which might instead develop in a bare frame solution.
On the other side, the sudden reduction of storey stiffness due to the damage of the infills can lead to the
formation of an unexpected soft storey mechanism, which, due to the interaction with the joint damage, can
occur not necessarily at the first floor level and independently by the regular or irregular distribution of the
infills along the elevation. Similarly, when investigating the response of 3-D frames under either uni-
directional or bi-directional earthquake input excitation, inelastic torsion mechanisms can occur.
In conclusion, it is worth recognizing that the high dispersion of the mechanical properties of the infills can
further increase the level of uncertainties in the expected performance if simplified probabilistic approach are
adopted

-i-
Acknowledgements

ACKNOWLEDGEMENTS

- ii -
Index

EVALUATION OF THE SEISMIC RESPONSE OF EXISTING R.C.


FRAME BUILDINGS WITH MASONRY INFILLS

INDEX

ABSTRACT ...................................................................................................................................................... I
ACKNOWLEDGEMENTS ............................................................................................................................II
INDEX ............................................................................................................................................................ III
LIST OF TABLES...........................................................................................................................................V
LIST OF FIGURES....................................................................................................................................... VI
1. INTRODUCTION ...................................................................................................................................1
2. STATEMENT OF THE PROBLEM .....................................................................................................4
3. LITERATURE REVIEW .....................................................................................................................11
4. MODELLING ISSUES .........................................................................................................................16
4.1. MODELLING OF STRUCTURAL ELEMENTS ........................................................................................17
4.1.1 Fiber Element Model ..................................................................................................................17
4.1.2 Lumped Plasticity Approach ......................................................................................................18
4.2. JOINT MODELLING ...........................................................................................................................20
4.2.1 Finite Element and Multi-Spring Models ...................................................................................20
4.2.2 Concentrated Plasticity Model....................................................................................................24
4.3. MODELLING OF MASONRY INFILLS ..................................................................................................26
4.3.1 Multi-Spring Models ..................................................................................................................27
4.3.2 Mechanical Properties of the Diagonal Strut..............................................................................29
5. VALIDATION OF THE ANALYTICAL MODEL............................................................................33
5.1. BEAM-COLUMN SUBASSEMBLIES ....................................................................................................34
5.1.1 Material properties......................................................................................................................34
5.1.2 Knee joint ...................................................................................................................................35
5.1.3 Interior joint ................................................................................................................................37
5.1.4 Exterior joint...............................................................................................................................39
5.2. SPECIMENS MODELLING ..................................................................................................................41
5.2.1 Structural elements .....................................................................................................................41
5.2.2 Joint elements .............................................................................................................................41
5.2.3 Joint hystetesis rule.....................................................................................................................42
5.2.4 Analytical – experimental comparison .......................................................................................46
5.3. THREE STOREY FRAME ....................................................................................................................50
5.3.1 Frame geometry and reinforcement details ................................................................................50

- iii -
Index

5.3.2 Material properties......................................................................................................................52


5.3.3 Test setup and loading history ....................................................................................................53
5.3.4 Description of the model ............................................................................................................55
5.3.5 Performed analyses.....................................................................................................................58
5.3.6 Analytical-experimental comparison..........................................................................................59
5.4. REINFORCED COCRETE INFILLED FRAMES.......................................................................................69
5.4.1 One storey infilled frames ..........................................................................................................69
5.4.2 Multi-storey three-dimensional infilled frames ..........................................................................74
6. NUMERICAL INVESTIGATION ON 2D FRAME SYSTEMS.......................................................77
6.1. THREE STOREY FRAME ....................................................................................................................78
6.1.1 Description of the infilled model ................................................................................................78
6.1.2 Pushover analyses of infilled frames ..........................................................................................80
6.1.3 Records used for the time history analyses.................................................................................84
6.1.4 Time history results ....................................................................................................................85
6.2. SIX STOREY FRAME .......................................................................................................................112
6.2.1 Description of the frame ...........................................................................................................113
6.2.2 Frame modeling ........................................................................................................................115
6.2.3 Pushover analysis......................................................................................................................116
6.2.4 Time history analyses ...............................................................................................................120
7. NUMERICAL INVESTIGATION ON 3D FRAME SYSTEMS.....................................................139
7.1. DESCRIPTION OF THE FRAMES .......................................................................................................140
7.1.1 Three storey frame ....................................................................................................................140
7.1.2 Six storey frame........................................................................................................................141
7.2. DESCRIPTION OF THE MODEL ........................................................................................................143
7.3. PUSHOVER ANALYSIS ....................................................................................................................144
7.3.1 Three storey frame ....................................................................................................................144
7.3.2 Six storey frame........................................................................................................................150
7.4. TIME HISTORY ANALYSIS ..............................................................................................................156
8. CONCLUSIONS ..................................................................................................................................171
9. REFERENCES ....................................................................................................................................174

- iv -
List of Tables

LIST OF TABLES

Table 4.1: Parameters of the equivalent diagonal strut model (Bertoldi, Decanini, Gavarini, 1993)................31
Table 5.1: Specimen material properties (Pampanin, 2002)..............................................................................34
Table 5.2: Specimen L1 reinforcement (Pampanin, 2002)................................................................................35
Table 5.3: Specimen C2 reinforcement (Pampanin, 2002)................................................................................37
Table 5.4: Specimen T1 and T2 reinforcement (Pampanin, 2002)....................................................................39
Table 5.5: Parameters needed to define hysteresis rule adopted for joint members..........................................44
Table 5.6: Calibration of the hysteretic rule parameters for the beam-column subassemblies .........................45
Table 5.7: Concrete cylinder compressive strength values (Pampanin, 2002)..................................................52
Table 5.8: Mean steel strengths of longitudinal bars (Pampanin, 2002) ...........................................................52
Table 5.9: Gravity load distribution (Pampanin, 2002).....................................................................................54
Table 5.10: Interaction diagram values for the frame columns .........................................................................56
Table 5.11: Floor displacement ratio applied in the pushover analysis.............................................................58
Table 5.12: Ratio of storey forces applied in the pushover analysis .................................................................58
Table 6.1: Masonry mechanical properties: mean value and c.o.v....................................................................79
Table 6.2: Record used for the time history analyses........................................................................................84
Table 6.3: Scale factors for the principal physical quantities............................................................................85
Table 6.4: Displacement and interstorey drift: mean values and standard deviation of ten analyses on
the bare frame.....................................................................................................................................86
Table 6.5: Displacement: mean values and standard deviation of ten analyses on the partially infilled
frame ..................................................................................................................................................94
Table 6.6: Interstorey drift: mean values and standard deviation of ten analyses on the partially infilled
frame ..................................................................................................................................................94
Table 6.7: Displacement: mean values and standard deviation of ten analyses on the uniformly infilled
frame ................................................................................................................................................103
Table 6.8: Interstorey drift: mean values and standard deviation of ten analyses on the uniformly
infilled frame....................................................................................................................................103
Table 6.9: Column reinforcement summary table ...........................................................................................114
Table 6.10: Displacement and interstorey drift: mean values and standard deviation of ten analyses on
the bare frame...................................................................................................................................121
Table 6.11: Displacement: mean values and standard deviation of ten analyses on the partially infilled
frame ................................................................................................................................................127
Table 6.12: Interstorey drift: mean values and standard deviation of ten analyses on the partially
infilled frame....................................................................................................................................128
Table 6.13: Displacement: mean values and standard deviation of ten analyses on the uniformly infilled
frame ................................................................................................................................................133
Table 6.14: Interstorey drift: mean values and standard deviation of ten analyses on the uniformly
infilled frame....................................................................................................................................134

-v-
List of Figures

LIST OF FIGURES

Figure 2.1: Damage on corner joint in 1999 Izmit, Turckey earthquake (NISEE, University of
California, Berkeley)...........................................................................................................................5
Figure 2.2: Development of a failure mechanism typical of poorly detailed exterior beam-column
joints (Pampanin, 2002) ......................................................................................................................6
Figure 2.3: Soft-storey mechanism observed in 1999 Izmit Earthquake, Turkey (Baletta, 2002). ....................7
Figure 2.4: Detailing of pre-1970s concrete structure in the Mediterranean countries (Italy) (Pampanin,
2003). ..................................................................................................................................................8
Figure 2.5: Damage resulted from column-infill wall interaction, Izmit, Turkey 1999 (NISEE,
University of California, Berkeley).....................................................................................................9
Figure 2.6: Failure of an exterior masonry infill panel during the 2002 Molise-Puglia Earthquake,
Italy. ..................................................................................................................................................10
Figure 3.1: Test-frame: elevation view and reinforcement layout (Calvi et al., 2002) ....................................12
Figure 3.2: Test-frame specimen: geometrical and mechanical characteristics (Colangelo, 1999) .................13
Figure 3.3: Layout of frame speciemen (Negro et al., 1995) ...........................................................................13
Figure 4.1: Beam element and corresponding fiber scheme (Monti and Spacone, 2000)................................17
Figure 4.2: Lumped plasticity beam element ...................................................................................................18
Figure 4.3: Hysteresis rules: a) Modified Takeda (Otani,1974); b) Fukada (Fukada,1969) ............................19
Figure 4.4: Interaction surface for a reinforced concrete column (Carr, 2004)................................................19
Figure 4.5: The effect of shear hinges on the sway mechanism of a frame (Trowland, 2003) ........................20
Figure 4.6: Finite element model (Nagai, 1996) ..............................................................................................21
Figure 4.7: Model of test specimen (Eligehausen et al., 2006) ........................................................................21
Figure 4.8: Multi-spring model proposed by Youssef and Ghobarah (2001)...................................................22
Figure 4.9: Joint model proposed by Elmorsi, Kianoush and Tso (2000)........................................................23
Figure 4.10: Reinforced concrete beam-column joint model (Lowes et al., 2003)..........................................23
Figure 4.11: Lumped plasticity model for beam-column joints (Pampanin et al., 2002).................................24
Figure 4.12: pt-γ relationships proposed for exterior (a) and interior joints (b) ...............................................25
Figure 4.13: Modified joint model representation (Trowland, 2003)...............................................................26
Figure 4.14: Equivalent diagonal strut model ..................................................................................................27
Figure 4.15: Possible variations of the classical equivalent diagonal strut model ...........................................28
Figure 4.16: Sliding shear infill model (Leuchars and Scrivener, 1973)..........................................................29
Figure 4.17: Histeretic cycle (a) and backbone curve (b) of Crisafulli model (1997)......................................32
Figure 5.1: Specimen L1: geometric characteristics and steel bars layout (Pampanin, 2002) .........................35
Figure 5.2: Test setup of specimen L1 (Pampanin, 2002)................................................................................36
Figure 5.3: Specimen C2: geometric characteristics and steel bars layout (Pampanin, 2002) .........................37
Figure 5.4: Test setup of specimen C2 (Pampanin, 2002)................................................................................38
Figure 5.5: Specimen T1: geometric characteristics and steel bars layout (Pampanin, 2002) .........................39
Figure 5.6: Reinforcement layout of specimens T1 and T2 (Pampanin, 2002)................................................40
Figure 5.7: Test setup of specimen T1 and T2 (Pampanin, 2002)....................................................................40
Figure 5.8: Modified joint model representation (Trowland, 2003).................................................................41

- vi -
List of Figures

Figure 5.9: Pampanin histeretic rule (Carr, 2004): a) Option 1-reloading power factor; b) Option 2-
reloading slip factor ..........................................................................................................................43
Figure 5.10: Typical loading history (Pampanin, 2002)...................................................................................46
Figure 5.11: Analytical-experimental comparison for specimen L1 ................................................................47
Figure 5.12: Analytical-experimental comparison for specimen C2................................................................48
Figure 5.13: Analytical-experimental comparison for specimen T1 and T2....................................................49
Figure 5.14: Geometrical properties and reinforcement layout of the test-frame (Pampanin, 2003)...............51
Figure 5.15: Column lap splice at floor level (units in cm) (Pampanin, 2003) ................................................51
Figure 5.16: Test setup (Pampanin, 2003)........................................................................................................53
Figure 5.17: Lateral loading history and force distribution (Pampanin, 2002) ................................................54
Figure 5.18: Geometric dimensions of the frame model ..................................................................................55
Figure 5.19: Interaction diagrams of the frame columns..................................................................................56
Figure 5.20: Modified Takeda hysteresis rules(Otani,1974) ............................................................................57
Figure 5.21: Fukada hysteresis rule (Fukada, 1969) ........................................................................................57
Figure 5.22: Imposed displacement history for the cyclic analysis..................................................................59
Figure 5.23: Comparison between pushover curve (displacement control), ....................................................60
Figure 5.24: Force-controlled pushover: base shear – top drift curve and displacement profile (after
Pampanin, 2002) ...............................................................................................................................61
Figure 5.25: Crack pattern observed at 1.2% top drift (Pampanin, 2003)........................................................62
Figure 5.26: Frame damage photo report (Pampanin, 2002)............................................................................63
Figure 5.27: Damage pattern and curvature ductility at 1.2% top drift level (displacement controlled
pushover)...........................................................................................................................................64
Figure 5.28: Damage pattern and curvature ductility at 1.2% top drift level (force controlled pushover).......64
Figure 5.29: Analytical-experimental comparison of the global hysteretic behavior of the frame (after
Pampanin, 2002) ...............................................................................................................................65
Figure 5.30: Experimental and analytical equivalent viscous damping values (after Pampanin, 2002) ..........66
Figure 5.31: Test frame joint identification labels ...........................................................................................67
Figure 5.32: Analytical and experimental joint shear deformation (after Pampanin, 2002) ............................68
Figure 5.33: Test frames: a) L2 specimen; b) N1 specimen (Colangelo, 2003)...............................................70
Figure 5.34: Bricks adopted for the specimens L2 (left) and N1 (right) (Colangelo, 2003) ............................70
Figure 5.35: Analytical experimental comparison for specimen L2: a) Top drift time history; b) Base
shear time hystory (Baletta, 2002) ....................................................................................................72
Figure 5.36: Analytical experimental comparison for specimen N1: a) Top drift time history; b) Base
shear time hystory (Baletta, 2002) ....................................................................................................73
Figure 5.37: Plan and elevation view of the test frame (Negro, 1995).............................................................75
Figure 5.38: Test specimen (Negro, 1995) .......................................................................................................75
Figure 5.39: Analytical-experimental comparison of the uniformly infilled frame: first floor
displacement and base shear time history (Galli, 2003)....................................................................76
Figure 5.40: Analytical-experimental comparison of the partially infilled frame: first floor
displacement and base shear time history (Galli, 2003)....................................................................76
Figure 6.1: Three storey frame infills distribution: a) uniformly: b) partially..................................................78
Figure 6.2: Base shear-top drift curve of the uniformly infilled frame ............................................................80
Figure 6.3: Displacement profile of the uniformly infilled frame ....................................................................81
Figure 6.4: Base shear-top drift curve of the partially infilled frame ...............................................................82
Figure 6.5: Displacement profile of the partially infilled frame.......................................................................82
Figure 6.6: Comparison between the 3 storey frame pushover curves.............................................................83
Figure 6.7: Comparison between EC8 elastic spectrum (PGA = 0.3g; Soil type B) and mean response
spectrum of ten records .....................................................................................................................84
Figure 6.8: Time-history analysis on the 3-storey bare frame: interstorey drift and displacement profile ......86
Figure 6.9: Bare frame: time history analysis results (Loma Prieta, EQ4).......................................................87
Figure 6.10: Bare frame: time history analysis results (Northridge, EQ8).......................................................88
Figure 6.11: Bare frame: time history analysis results (Superstition Hills, EQ10) ..........................................89
Figure 6.12: Bare frame joint rotations ............................................................................................................90
Figure 6.13: Bare frame: elements contribution to the maximum interstorey drift (EQ4)...............................91
Figure 6.14: Bare frame: elements contribution to the maximum interstorey drift (EQ8)...............................92
Figure 6.15: Bare frame: elements contribution to the maximum interstorey drift (EQ10).............................93

- vii -
List of Figures

Figure 6.16: Time-history analysis on the 3-storey partially infilled frame: interstorey drift and
displacement profile (average of maximum and residual) ................................................................95
Figure 6.17: Partially infilled frame: time history analysis results (Loma prieta, EQ4) ..................................96
Figure 6.18: Partially infilled frame: time history analysis results (Northridge, EQ8) ....................................97
Figure 6.19: Partially infilled frame: time history analysis results (Superstition Hills, EQ10)........................98
Figure 6.20: Partially infilled frame joint rotations..........................................................................................99
Figure 6.21: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ4) ..........100
Figure 6.22: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ8) ..........101
Figure 6.23: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ10) ........102
Figure 6.24: Time-history analysis on the 3-storey uniformly infilled frame: interstorey drift and
displacement profile (average of maximum and residual) ..............................................................104
Figure 6.25: Uniformly infilled frame: time history analysis results (Loma Prieta, EQ4).............................105
Figure 6.26: Uniformly infilled frame: time history analysis results (Northridge, EQ8)...............................106
Figure 6.27: Uniformly infilled frame: time history analysis results (Superstition Hills, EQ10) ..................107
Figure 6.28: Uniformly infilled frame joint rotations.....................................................................................108
Figure 6.29: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ4) .......109
Figure 6.30: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ8) .......110
Figure 6.31: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ10) .....111
Figure 6.32: Geometric dimensions of the 6-storey frame .............................................................................112
Figure 6.33: Beam sections: geometric and mechanical characteristics.........................................................114
Figure 6.34: Column sections of the 6-storey frame ......................................................................................114
Figure 6.35: Pushover analysis on the 6-storey bare frame: Base shear - Top drift curve .............................116
Figure 6.36: Pushover analysis on the 6-storey bare frame: Displacements and Interstorey drift profiles
corresponding to top drift value of 0.5% and 1%. ..........................................................................117
Figure 6.37: Pushover analysis on the 6-storey partially infilled frame: Base shear - Top drift curve ..........118
Figure 6.38: Pushover analysis on the 6-storey partially infilled frame: Displacements and Interstorey
drift profiles corresponding to top drift value of 0.5% and 1%. .....................................................118
Figure 6.39: Pushover analysis on the 6-storey uniformly infilled frame: Base shear - Top drift curve .......119
Figure 6.40: Pushover analysis on the 6-storey uniformly infilled frame: Displacements and Interstorey
drift profiles corresponding to top drift value of 0.5% and 1%. .....................................................119
Figure 6.41: Comparison between the 6 storey frame pushover curves.........................................................120
Figure 6.42: Time-history analysis on the 6-storey bare frame: interstorey drift and displacement
profile (average of maximum and residual) ....................................................................................121
Figure 6.43: Bare frame: time history analysis results (Loma Prieta, EQ4)...................................................123
Figure 6.44: Bare frame: time history analysis results (Northridge, EQ8).....................................................124
Figure 6.45: Bare frame: time history analysis results (Superstition Hills, EQ10) ........................................125
Figure 6.46: Bare frame joint rotations ..........................................................................................................126
Figure 6.47: Time-history analysis on the 6-storey partially infilled frame: interstorey drift and
displacement profile (average of maximum and residual) ..............................................................127
Figure 6.48: Partially infilled frame: time history analysis results (Loma Prieta, EQ4) ................................129
Figure 6.49: Partially infilled frame: time history analysis results (Northridge, EQ8) ..................................130
Figure 6.50: Partially infilled frame: time history analysis results (Superstition Hills, EQ10)......................131
Figure 6.51: Partially infilled frame joint rotations........................................................................................132
Figure 6.52: Time-history analysis on the 6-storey uniformly infilled frame: interstorey drift and
displacement profile (average of maximum and residual) ..............................................................133
Figure 6.53: Uniform infilled frame: time history analysis results (Loma Prieta, EQ4)................................135
Figure 6.54: Uniform infilled frame: time history analysis results (Northridge, EQ8) ..................................136
Figure 6.55: Uniform infilled frame: time history analysis results (Superstition Hills, EQ10) .....................137
Figure 6.56: Uniformly infilled frame joint rotations.....................................................................................138
Figure 7.1: Three storey 3D frame: plan view and beam section properties ..................................................140
Figure 7.2: Three storey 3D frame: infills layout...........................................................................................141
Figure 7.3: Six storey frame plan view...........................................................................................................142
Figure 7.4: Transversal beam mechanical and geometrical properties...........................................................142
Figure 7.5: Six storey frame: infills layout.....................................................................................................143
Figure 7.6: Three dimensional interaction surface adopted for columns (Carr, 2004)...................................143
Figure 7.7: 3D bare frame pushover: X direction...........................................................................................144

- viii -
List of Figures

Figure 7.8: 3D bare frame pushover: Z direction ...........................................................................................145


Figure 7.9: 3D bare frame pushover: displacement profiles...........................................................................146
Figure 7.10: 3D partially infilled frame pushover: X direction......................................................................146
Figure 7.11: 3D partially infilled frame pushover: Z direction ......................................................................147
Figure 7.12: 3D partially infilled frame pushover: displacement profiles.....................................................147
Figure 7.13: 3D uniformly infilled frame pushover: X direction ...................................................................148
Figure 7.14: 3D uniformly infilled frame pushover: Z direction....................................................................148
Figure 7.15: 3D uniformly infilled frame pushover: displacement profiles ..................................................149
Figure 7.16: 3D bare frame pushover: X direction.........................................................................................150
Figure 7.17: 3D bare frame pushover: Z direction .........................................................................................151
Figure 7.18: 3D bare frame pushover: displacement profiles.........................................................................151
Figure 7.19: 3D partially infilled frame pushover: X direction......................................................................152
Figure 7.20: 3D partially infilled frame pushover: Z direction ......................................................................153
Figure 7.21: 3D partially infilled frame pushover: displacement profiles.....................................................153
Figure 7.22: 3D uniformly infilled frame pushover: X direction ...................................................................154
Figure 7.23: 3D uniformly infilled frame pushover: Z direction....................................................................155
Figure 7.24: 3D uniformly infilled frame pushover: displacement profiles ...................................................155
Figure 7.25: Three storey frame model and reference system........................................................................156
Figure 7.26: Bare frame: maximum displacements........................................................................................157
Figure 7.27: Bare frame: maximum interstorey drift .....................................................................................158
Figure 7.28: Bare frame: residual displacements ...........................................................................................159
Figure 7.29: Bare frame: residual interstorey drift .........................................................................................160
Figure 7.30: Partially infilled frame: maximum displacements .....................................................................161
Figure 7.31: Partially infilled frame: maximum interstorey drift ...................................................................162
Figure 7.32: Partially infilled frame: residual displacements.........................................................................163
Figure 7.33: Partially infilled frame: residual interstorey drift.......................................................................164
Figure 7.34: Uniformly infilled frame: maximum displacements ..................................................................165
Figure 7.35: Uniformly infilled frame: maximum interstorey drift................................................................166
Figure 7.36: Uniformly infilled frame: residual displacements......................................................................167
Figure 7.37: Uniformly infilled frame: residual interstorey drift ...................................................................168
Figure 7.38: 3D time history analysis: maximum floor diaphragm rotations.................................................170

- ix -
Chapter 1 – Introduction

1. INTRODUCTION

In the recent years the assessment of the seismic response of existing buildings is becoming a
widely investigated topic. Special attention has been given to the vulnerability of reinforced
concrete frames designed for gravity loads only, that was common practice in several
Mediterranean countries before the introduction of specific seismic design provisions. Most of these
structures have been typically designed before the 1970s when "capacity design" philosophy was
not yet widely introduced in seismic design codes. Current seismic design procedures around the
world have advanced significantly when compared to the time when those reinforced concrete
structures were built. The main advantages are the understanding of the post-elastic seismic
behavior of the structures and the improvement in the structural detailing to enhance the ductility
capacity of the structural components. Therefore, many existing reinforced concrete structures
designed and constructed before '70s may present typical deficiencies such as: (a) use of plain round
bars as longitudinal reinforcement, (b) inadequate anchorage of beam longitudinal reinforcement in
the column, (c) lack of joint transverse reinforcement and (d) lapped splices located just above joint.
The importance of the seismic assessment of existing reinforced concrete structures has in a
relatively recent past received more and more attention following the high level of damage and
socio-economical consequences observed as a result of recent severe earthquake events. Several
recent earthquakes such as Hyogo-ken Nanbu Earthquake in Japan (1995), Chi-Chi Earthquake in
Taiwan (1999), Izmit Earthquake in Turkey (1999) and more recent seismic events in Southern
Italy, caused heavy damage and collapses to an excessively high number of existing reinforced
concrete structures designed to outdated codes.
On the other hand, most of the reinforced concrete structures built recently and designed according
to modern seismic code provisions, did not suffer significant damage, proving the current
knowledge on seismic design to be relatively advanced and the correspondent design guidelines
adequate in limiting the disastrous consequences observed in the past.

-1-
Chapter 1 – Introduction

Seismic assessment is the first step within the retrofit strategy to reduce the seismic risk. A good
understanding of the weak point of a structure under seismic loading could allow to complement
and design the most appropriate retrofit solution to reduce the seismic vulnerability.
Some of the seismic assessment procedures are already advanced using capacity design philosophy
and taking into account the global structural behavior in the post-elastic range. Experimental
research to investigate the possible seismic behavior of pre-1970s reinforced concrete structures
have been carried out in the past. However, there is still a need of further investigation to
understand more the response of R.C. buildings under seismic loading.
Scope of this work is to investigate the seismic response or reinforced concrete frame buildings and
the influence of masonry infills on the global structural response. To better understand the problem
is important to perform a large number of parametric analyses taking into account various structural
layout and masonry infills distributions. The need to perform a large number of parametrical
analyses brings to light the importance of developing a numerical model which ensures a good
compromise between simplicity, especially in the computation and elaboration of results, and
sufficient refinement to adequately capture the different mechanisms that influence the structural
response. In the present work the model used for the numerical analyses is based on the lumped
plasticity approach, with members having the inelastic regions located at their ends. An interesting
feature is the model refinement obtained with the introduction of non-linear rotational springs to
represent the effect of beam-column joint shear deformation on the overall response of the frame.
Another objective of this work is to investigate the influence of the presence of masonry infills on
the seismic behavior of existing R.C. frame buildings. Due to the high variability of mechanical
properties of masonry and to the uncertainty in the determination of parameters necessary for the
modeling of masonry struts, the approach adopted was based on the identification of an upper and
lower bound in the infills characteristics trying to obtain a reasonable range of results representing a
wide number of possible situation that can be met in the practice. In the analyses performed in this
work the lower bound was represented by a weak infill constituted by a single leaf panel and the
upper bound has been identified in a stronger double leaf infill panel.
The final objective of this work is the investigation of the three-dimensional response of the frame
structures and the influence that structural configuration, infills layout and direction of seismic
action can have on the development of torsional effects on the frame behavior.

All the numerical analyses have been run using the Inelastic Dynamic Analysis Program
RUAUMOKO (Carr, 2004).

-2-
Chapter 1 – Introduction

Thesis outline

This thesis consists of eight chapters that, a part from this first introductive chapter, can be briefly
resumed as follows.
Chapter 2 presents the typical seismic deficiencies of pre-1970s concrete frame structures, with and
without masonry infills. This review clarify the importance of understanding the actual local
behavior of concrete structure components for conducting the seismic assessment of structures
designed for gravity loads only.
Chapter 3 gives a review of previous researches carried out by different authors on the topic, to
identify what has been done and what still needs to be done for a more comprehensive
understanding of the seismic performance of existing frame buildings.
Chapter 4 presents an overview of different modeling approaches existing in literature for various
structural and non-structural components which constitute the building. Particular attention is
focused on the modeling solutions adopted in the present work such as equivalent diagonal struts
used to represent the masonry infills and rotational springs which have been adopted to model the
beam-column joints.
Chapter 5 describes the validation of the analytical model carried out by comparison of numerical
and experimental results. The validation of joint springs had been performed comparing the
numerical results with results obtained from the experimental tests on beam-column subassemblies.
The results of the global frame model has been compared with the results of cyclic tests performed
on a 2/3 scaled three storey reinforced concrete bare frame. Finally the validation of equivalent
diagonal struts are carried base on experimental results obtained from tests on 2D and 3D reinforced
concrete frame structures.
Chapter 6 presents the results of numerical investigations carried out on two dimensional frame
structures with different structural configurations and infills layout. At first pushover analyses were
performed to better understand the global behavior on the structures subjected to seismic loading.
Then the structural response is investigated by mean of non-linear time history analyses on the
structural models.
Chapter 7 presents results of numerical analyses performed on three dimensional frames
considering various structural configurations and infills disposition. It has also been investigated the
influence of the direction and inclination of seismic loading on the final response of the structure
and the possible development of torsional effects affecting the frame behavior.
Chapter 8 summarize the conclusions reached in this work and gives suggestions for further
development and future research investigations.

-3-
Chapter 2 – Statement of the problem

Equation Section 2

2. STATEMENT OF THE PROBLEM

The construction and design practice of reinforced concrete structures have significantly advanced
around the world since the 1970's, mainly in the understanding of the seismic hazard. Current
seismic design provisions require a structure to have adequate reinforcement detailing to provide an
adequate ductile behavior necessary to resist a targeted level earthquake (i.e. a given return period).
Up to the 70's, most of the structures were not designed using 'capacity design' principles and the
seismic detailing was poor if compared to those currently implemented in more recent design codes.
This chapter reviews the typical detailing and the deficiencies of pre-1970s reinforced concrete
structures.
As mentioned most of the pre-1970s reinforced concrete structures were not designed with capacity
design philosophy, which means an undesirable seismic mechanism such as a soft-storey
mechanism might occur. It is found that columns were designed to provide strength only to restrain
the lateral load applied on the structure , not on the possible load coming from the beams. This will
lead to weak column-strong beam mechanism which have disastrous outcomes.
Due to the absence of 'capacity design' philosophy before the 1970s, the lack of ductility has been
acknowledged as one of the main reason of the unsatisfactory seismic performance of the reinforced
concrete structures designed during that period and which is accentuated by poor reinforcement
detailing. Typical structural deficiencies found in these buildings are:

• lack of appropriate confinement through transverse reinforcement in the plastic hinge


regions;

• lack or even absence of transverse reinforcement in the joint core;

• moment capacity of the column is lower then that of beams;

• use of plain round bars instead of deformed bars and inadequate reinforcement
anchorage;

-4-
Chapter 2 – Statement of the problem

• lap splices located in potential member plastic hinge regions;

• low strength and poor quality material when compared to present practice;

• presence of masonry infill walls with complex interaction with bare frame;
Most of the reinforced concrete structures designed in the pre-1970s exhibit a lack of shear
reinforcement in the joint region. Sometimes, one stirrup only or no shear reinforcement at all was
provided in the joint core. Without adequate shear reinforcement in the joint, the concrete has to
resist all the forces from the beam and the column passing through the joint, which may lead to a
joint shear failure mechanism. High diagonal compressive and tensile stresses occur in the joint as
the result of the shear forces. The cracking in the concrete develops according to the tensile stresses
and it will lead to a large shear deformation of the joint. The term 'shear hinge' (Pampanin, 2003)
has been used to describe this mechanism, alternative and dual to a typical flexural plastic hinge.

Figure 2.1: Damage on corner joint in 1999 Izmit, Turckey earthquake (NISEE, University of California,
Berkeley)

-5-
Chapter 2 – Statement of the problem

Figure 2.2: Development of a failure mechanism typical of poorly detailed exterior beam-column joints
(Pampanin, 2002)

In most of the pre-1970s reinforced concrete structures, the columns were mainly designed for
gravity load only or to resist the bending moment from the low level of lateral forces specified by
the code (typically a small portion of the building weight) without consideration on the relative
hierarchy of strength with the beam moment capacity. This may cause the column to be weaker than
the beams, and possibly resulting in the column side sway mechanism (i.e. soft-storey mechanism),
rather then the preferred beam sway mechanism.

-6-
Chapter 2 – Statement of the problem

Figure 2.3: Soft-storey mechanism observed in 1999 Izmit Earthquake, Turkey (Baletta, 2002).

In addition, as mentioned, plain round bars with hook-end were widely used for longitudinal
reinforcement in the reinforced concrete structures before the 1970s. The performance of a structure
using plain round bars with end hooks as longitudinal reinforcement can result to a very poor
behavior under reversed cyclic inelastic loading due to the lack of bond strength between the steel
and the concrete which leads to bar slipping and high global deformation of the system. Anchorage
provided by hook-end bars was likely not to be sufficient to prevent the bars from slipping and the
concentration of strut and compression force at the hook can lead to a peculiar damage and failure
mechanism due to the expulsion of a 'concrete wedge' (Pampanin, 2003).
The location of lap splices of the longitudinal reinforcement is another seismic inadequacy typically
found in pre-1970s reinforced concrete structures. Lapped splices are usually located in the plastic
hinge regions of the beams and just above the beam-column joint area, where the maximum
moment develop. This will lead to inadequate local ductility for the beams and columns.

-7-
Chapter 2 – Statement of the problem

Figure 2.4: Detailing of pre-1970s concrete structure in the Mediterranean countries (Italy) (Pampanin, 2003).

The presence of infills, if on one hand can increase the structural resistance of a reinforced concrete
frame, on the other hand could also provide controversial effects to the structural capacity as well as
increase the seismic demand (Crisafulli, 1997). Infill walls provide additional stiffness to the
structure reducing the deformation demand. On the other side, the increment of structure's lateral
stiffness will reduce its fundamental period; therefore there will be an increase in seismic action. If
the use of infills is not distributed evenly in the frame, it can alter the structural mechanism. For

-8-
Chapter 2 – Statement of the problem

example in 1999 Chi-Chi Earthquake in Taiwan or in the 1999 Izmit Earthquake in Turkey, a lot of
buildings show a soft-storey mechanism due to the irregular distribution of infills. Typical practice
is to use infill panels in the higher storey leaving the ground storey free from infill walls. Another
source of irregularity is the large scatter that characterizes the mechanical properties of masonry
infills. The concomitance of these different aspects can lead to unexpected or peculiar effects on the
global response of infilled frames if compared with bare frames, like the formation of soft storey
mechanisms of the shear failure of columns due to the interaction between the masonry panel and
the surrounding frame.

Figure 2.5: Damage resulted from column-infill wall interaction, Izmit, Turkey 1999 (NISEE, University of
California, Berkeley).

Typically the masonry infill panels can show three different failure mechanism when subjected to
cyclic loading. When the shear stresses present in the panel are significant if compared with the
compression stresses perpendicular to the mortar layers, is likely that a shear failure mechanism will
develop in the infill. This mechanism is characterized by the formation of typical "stepped" cracks
running along the diagonal of the wall panel. Another possible failure mechanism is that associated
with the diagonal traction. This situation occurs when the principal tensile stress in the brick
overcomes the tensile resistance, causing the development of diagonal cracks starting from the
center of the panel towards the corners, along the panel diagonal. The compression failure
mechanism, instead, is likely to develop when the infill panel is surrounded by a flexible frame. The
lateral deformation of the frame can produce a reduction in the contact length between the infill and
the columns, causing a dramatic increase of compressive stress in the panel corners, leading to the
masonry failure. It is worth noting however that often the infill panels show hybrid failures in which
different mechanisms can coexist.

-9-
Chapter 2 – Statement of the problem

Figure 2.6: Failure of an exterior masonry infill panel during the 2002 Molise-Puglia Earthquake, Italy.

- 10 -
Chapter 3 – Literature Review

Equation Section (Next)

3. LITERATURE REVIEW

The topic of seismic assessment of existing reinforced concrete frame structures and the
investigation of seismic response of infilled frame systems is a topic that recently has gathered the
interest of different researchers. Studies have been carried out both on existing under designed
buildings and on R.C. infilled frames designed following the recent seismic code provisions. In this
chapter an overview of a series of studies performed by different authors is presented.

The seismic response of existing reinforced concrete frame structures designed for gravity loads
only has been investigated by Calvi, Magenes and Pampanin (2002). A quasi-static cyclic
experimental test on a three storey frame system, 2/3 scaled, were performed at the Laboratory of
Structural Mechanics of the University of Pavia. The frame tested shows the structural inadequacies
typical of pre-seismic code provisions such as: plain round bars, inadequate reinforcement detailing
and absence of any capacity design principle that result in brittle local and global failure
mechanisms. Particularly critical joint damage, with no alternative sources for gravity-load bearing
capacity, was observed in the exterior joints. On the basis of the observed global frame response,
the concept of "shear hinge", due to joint damage, is introduced by the authors as an alternative to
flexural plastic hinge in structural members.

- 11 -
Chapter 3 – Literature Review

Figure 3.1: Test-frame: elevation view and reinforcement layout (Calvi et al., 2002)

Colangelo (1999) presents the results obtained from experimental tests performed at the DISAT
laboratory of the University of L'Aquila on RC frames with weak columns, filled in with hollow-
brick masonry. The work is mainly based on pseudo-dynamic tests carried out on single -storey
single -bay half-scale specimens. The response exhibited by bare and infilled frames is compared
and interpreted, stressing the influence of the infill on stiffness, strength, dissipation capacity (by
dynamic linear equivalent parameters too), cyclic deterioration, and failure mechanism. Colangelo
underlines the beneficial effect of masonry infills on the elastic response of existing reinforced
concrete frames. In particular the stiffness increases of one order of magnitude while the resistance
doubles respect to a bare frame. In the post-elastic range, instead, the presence of infills panel and
their rapid strength degradation can lead to the development of undesirable failure mechanisms.

- 12 -
Chapter 3 – Literature Review

Figure 3.2: Test-frame specimen: geometrical and mechanical characteristics (Colangelo, 1999)

Negro et al. (1995) performed a series of pseudo-dynamic tests on a three dimensional, full-scale
four-storey reinforced concrete frame designed in accordance with Eurocode 8 provisions. Three
different specimens have been considered: a bare frame, a uniform infilled frame and a partial
infilled frame without panels at the ground level. The results of pseudo-dynamic tests are presented
and in conclusion some analytical results obtained from numerical analyses are shown and
compared with experimental results.

D ire c tio n o f
te s tin g

Figure 3.3: Layout of frame speciemen (Negro et al., 1995)

- 13 -
Chapter 3 – Literature Review

Mosalam, White and Gergely carried out a work consisting of a first experimental phase and a
following phase focused on modeling and numerical investigation. A series of three reports present
the work carried out (Mosalam et al., 1997). The first report presents the results of cyclic test
performed on five 1/4 scaled infilled frame specimens, one of them is a single storey single bay
frame while the other four are single storey two bays frames. The structure is constituted of a steel
frame designed for gravity loads only. and the infill panels are made of hollow concrete blocks. The
parameters investigated during the experimental tests were the effect of different ratios of block to
mortar resistance, the presence and distribution of openings in the panel.
The second report presents the results of pseudo-dynamic tests performed on a two storey two bays
infilled frame characterized by an opening in the upper storey infill panels. The specimen has been
subjected to a sequence of accelerograms with increasing level of intensity. The crack pattern at
increasing level of peak ground acceleration has been investigated and a system of equivalent struts
representing the stress distribution is individuated in the central region of the panel.
The third report treats the modeling of infilled frames at different levels of refinement. The first
approach considers dimensionless elements to represent the mortar layers and a new constitutive
law has been proposed. The second approach, less refined but more computationally effective, is
based on masonry homogenization. After a numerical validation, the simplified model has been
used to derive the fragility curves of bare and infilled frames designed for gravity loads only.

Colombo, Negro & Verzeletti (1998) and Fardis, Bousias & Panagiotakos (1998) present
experimental results obtained from shake table tests, performed on a 2-storey frame considering at
first the bare frame configuration and then an infilled frame with irregular distribution of panels in
plant, and from pseudo-dynamic test on a three storey frame with different infills distribution along
building height. The specimen are full scale reinforced concrete frames infilled with masonry
hollow blocks. The study was finalized to the evaluation of the Eurocode 8 prescriptions regarding
the three-dimensional analysis of irregular structures

Cosenza, Manfredi and Verderame (2002) face the problem of seismic assessment of R.C. frames
designed for gravity loads only. A summary of models that permit the analysis of the non-linear
behavior of RC structures is discussed. An innovative numerical model is presented which takes
into account the most important mechanical phenomena affecting the non-linear behavior of the RC
frames. In conclusion, the influence of different strength and deformation sources on the global
behavior of existing buildings is studied and the needed capabilities of the numerical models are
underlined.

- 14 -
Chapter 3 – Literature Review

Attention to the theme of code prescriptions has been given by Fardis (1997), that on the basis of
extended numerical analyses performed using a global model for masonry infills, proposes some
modifications to the EC8 code provisions. The report by Fardis presents a series of local and global
models, validated comparing the numerical results with the results of pseudo-dynamic test
performed on a four storey, full scale infilled R.C. frame at the ELSA Laboratory at Ispra. A further
scope of the work was the investigation of the out-of-plane behavior of the infill panels. Some
experimental test results (monotonic and on shaking table) are reported together with the analytical
results of a proposed non-linear model which describes the in-plane and out-of-plane behavior of
infills

- 15 -
Chapter 4 – Modeling issues

Equation Section 4

4. MODELLING ISSUES

When the seismic response of a structure is investigated, it is essential to define an adequate


numerical model to describe the various aspect that contribute to the global structural behavior.
Several numerical models are available in literature depending on the refinement level that one
wants to achieve. Dealing with the vulnerability assessment of existing reinforced concrete frame
structure, with and without infills, both simplicity and reliability were targeted as fundamental
properties of the numerical model. It is considered inappropriate the use of a excessively refined
model when dealing with a topic characterized by several uncertainties in the geometrical and
mechanical properties of the structure object of study.
Furthermore in the last few decades a series of experimental tests performed on R.C. frame
buildings characterized by structural deficiencies typical of older design practice were performed,
resulting in a quite extended amount of data that can be used to better calibrate and refine simplified
numerical models.
In this chapter a brief review of the principal existing modeling solutions, for the various structural
and non-structural elements, is presented and they are compared with the model adopted in the
present work, based on the concentrated plasticity approach.

- 16 -
Chapter 4 – Modeling issues

4.1. MODELLING OF STRUCTURAL ELEMENTS

Various levels of refinement are available to describe the in-elastic behavior of linear structural
elements such as beams and columns. In particular the two most common approaches are the fiber
element model and the lumped plasticity model. In the following paragraphs a brief description of
these two types is presented.

4.1.1 Fiber Element Model

Each beam and column is divided into a finite number of elements depending on the variation of
geometric or mechanical properties. In this way is possible to take into account the different amount
of reinforcement along the element.
The sections of the element are subdivided into areas, called fibers, whose number and dimensions
vary depending on the level of precision one wants to achieve. After defining the materials
constitutive laws and the reinforcement distribution in the element section, the integration of
flexibility along the element length allows the determination of the stiffness matrix by inverting the
flexibility matrix. In this way is possible to describe the in-elastic behavior of the element governed
by flexure. It is to notice that this approach in the determination of the plasticity of the structural
member doesn't need the preliminary definition of the plastic hinge length that is a source of
uncertainties in the solution.
With this approach is also possible to describe slip phenomena occurring between concrete and steel
reinforcement by introducing an adequate stress-slip relationship. In figure 4.1 a structural member
and the corresponding fiber element schematization, proposed by Monti and Spacone (2000) is
shown.

Figure 4.1: Beam element and corresponding fiber scheme (Monti and Spacone, 2000)

- 17 -
Chapter 4 – Modeling issues

The fiber element approach can describe very precisely the various mechanisms influencing the
non-linear behavior of structural elements but to obtain reliable results a good level of knowledge of
structural properties (mechanical and geometrical) is needed. Often this is not the case when the
objective of the study are existing buildings for which high level of uncertainties usually exists.
Furthermore, due to the high refinement of the model, long time for the integration of the solution is
usually needed.

4.1.2 Lumped Plasticity Approach

The lumped plasticity approach is a good compromise between accuracy and simplicity. Using a
lumped plasticity model the parts of the member which are likely to undergo plastic deformations
have to be identified through a preliminary analysis. For frame structures this is a simple process
because plastic hinges generally form at the ends of the members. Beams and columns are modeled
by mono-dimensional elastic elements with in-elastic behavior concentrated at the ends in plastic
hinge regions (Giberson one component beam model) and defined by appropriate moment-curvature
hysteresis rules available in RUAUMOKO.

Elastic Element
Elemento elastico

Molle rotazionali
Plastic Hinges

Figure 4.2: Lumped plasticity beam element

The plastic hinge length has to be evaluated before the analysis using one of the several formula
present in literature. In the present work the following relationship has been used (Priestley et al.,
1996):
L
L p = 0.08 + 0.022 f yl dbl (4.1)
2
where Lp is the plastic hinge length, L is the element length, fyl is the yielding strength of
longitudinal reinforcement bars and dbl is the diameter of steel bars.
Typical hysteresis rules with bi-linear or tri-linear monotonic branch has been used to describe the
cyclic behavior of structural elements.

- 18 -
Chapter 4 – Modeling issues

a) b)

Figure 4.3: Hysteresis rules: a) Modified Takeda (Otani,1974); b) Fukada (Fukada,1969)

To take into account the effect of axial load variation on the capacity of the column elements an M-
N interaction diagram can be defined; furthermore strength degradation curves, function of number
of cycles or ductility demand, can be associated to the chosen hysteretic rule to consider the
increasing loss of strength in elements which experience in-elastic deformations.
When the three dimensional seismic response of a building wants to be investigated is important to
represent adequately the effective behavior of columns when subjected to bi-axial flexure together
with axial load variation. The RUAUMOKO program (Carr, 2004) allows to assign a three
dimensional interaction surface to the reinforced concrete column members with the possibility of
choosing between a linear or elliptic Mz – My interaction, see figure 4.4.

Figure 4.4: Interaction surface for a reinforced concrete column (Carr, 2004)

- 19 -
Chapter 4 – Modeling issues

This modeling approach allows to describe well the mechanisms governed by flexure without the
possibility of considering directly the occurrence of shear failures in the members, that should be
checked separately after the analysis.

4.2. JOINT MODELLING

An adequate modeling of the joint panel zone non-linear behavior is a critical basis for a correct
evaluation of the seismic response of the whole system. As shown by experimental tests on R.C.
frames designed for vertical loads only (Calvi, 2002), the damage occurring in the beam-column
joint can result in a modification of the deformed shape of the structure subjected to lateral loading,
with a spread of the inter-storey drift over several stories (Figure 4.5).
As for structural members, is possible to achieve different levels of detailing and complexity in
modeling the non-linear behavior of beam-columns joints. The most complete option available is to
use three dimensional non-linear finite element models. Less refined techniques can be chosen
based on the use of spring elements such as multi-spring models or lumped plasticity models which
vary depending on the number and characteristics of the element used to represent the joint.
In the following paragraphs a brief overview of finite element and multi-spring model if presented,
followed by a more accurate description of the lumped plasticity approach used in the present study.

Figure 4.5: The effect of shear hinges on the sway mechanism of a frame (Trowland, 2003)

4.2.1 Finite Element and Multi-Spring Models

Among finite elements models proposed, Nagai, Kashiwazaki and Noguchi (1996) used three
dimensional non-linear finite elements to model a high strength concrete joint subjected to biaxial
monotonic loading. The inelastic behavior of interior wide column joints subjected to uniaxial

- 20 -
Chapter 4 – Modeling issues

loading has been investigated by Bing, Yiming and Tso-Chien (2003) using two dimensional non-
linear finite elements. In figure 4.6 a sketch of the Nagai three dimensional model is shown.

Figure 4.6: Finite element model (Nagai, 1996)

A finite element approach specially developed for detailed modeling of fracture in quasi brittle
materials has been recently proposed by Eligehausen et al. (2006). The microplane material model
with relaxed kinematic constraint is used for the concrete and particular interest has been addressed
to the proper modeling of the behavior of smooth reinforcement with hooked ends as well as to the
accurate representation of brittle shear failure modes in joint. The microplane model is a three-
dimensional, macroscopic model in which the material is characterized by uniaxial relations
between the stress and strain components on planes of various orientations called "microplanes".
The discrete bond model implemented consists of a one-dimensional finite element with a realistic
bond-slip relationship. Figure 4.7 shows the beam-column joint model used for the validation.

Figure 4.7: Model of test specimen (Eligehausen et al., 2006)

- 21 -
Chapter 4 – Modeling issues

In the multi-spring model proposed by Youssef and Ghobarah (2001) the joint region is modeled by
four rigid elements that include the node and elastic members represent beams and columns
convergent into the joint. The model is completed by springs describing the concrete and steel
cyclic behavior to which the constitutive laws of the materials are associated. The shear deformation
of the joint region is modeled with two springs (shear spring) placed along the diagonals of the
node panel. Figure 4.8 shows a scheme of the multi-spring beam column joint model described
above.

Figure 4.8: Multi-spring model proposed by Youssef and Ghobarah (2001)

A different model for the joint region has been proposed by Elmorsi et al. (2000). In this approach
beams and columns are described by elastic elements and are connected to the joint through the
interposition of non-linear transitional elements. The effective node panel region is modeled with
another element constituted by 10 joints (Figure 4.9). This model allows to describe the material
behaviors with the introduction of the stress-strain relationships of steel and concrete. Concrete is
defined by two different relationships defining the pre and post cracking behavior. Longitudinal
reinforcing steel bars are modeled with non-linear elements placed along the upper and lower sides
of the joint panel. Furthermore this model allows the introduction of a "bond-slip element" to
represent the slipping of steel bars that in some cases plays a basic role in the seismic response of
under-designed reinforced concrete frames.

- 22 -
Chapter 4 – Modeling issues

Elastic column
element

-
Non-linear
transitional element

Elastic beam
element

Joint element

Figure 4.9: Joint model proposed by Elmorsi, Kianoush and Tso (2000)

Recently a multi-spring model for beam-column joints has been proposed by Lowes and Altoontash
(2003). The proposed model represents the basic mechanisms which govern the joint behavior such
as shear failure of joint panel and anchorage failure of beam and column longitudinal reinforcement
passing through the joint region. The model is constituted of eight bar-slip springs, four interface
shear springs and one shear-panel component. Different constitutive relationships, function of joint
geometry, material properties and reinforcement layout, can be associated to the various
components of the model in order to represent the non-linear joint response. Figure 4.10 shows a
scheme of the proposed joint model.

Figure 4.10: Reinforced concrete beam-column joint model (Lowes et al., 2003)

- 23 -
Chapter 4 – Modeling issues

4.2.2 Concentrated Plasticity Model

A simple model has been proposed by Pampanin et al. (2002), it consists of a non-linear rotational
spring that permits to model the relative rotation between beams and columns converging into the
node and to describe the post-cracking shear deformation of the joint panel. Beam and column
elements are modeled as one dimensional elements with lumped plasticity in the end sections with
an associated moment-curvature relationships defined by a section analysis. The effect of moment-
axial load interaction are taken into account for columns. To represent the real geometric
dimensions of the joint panel region, rigid elements are used to connect the beam and column
members to the rotational spring.

CLOSE UP VIEW
Panel Zone Region

RigidLink
element
rigidi

Molla rotazionale
Rotational spring

Figure 4.11: Lumped plasticity model for beam-column joints (Pampanin et al., 2002)

The definition of the moment-rotation relationship of the rotational spring is based on the results of
experimental tests performed at the Department of Structural Mechanics of the University of Pavia
(Pampanin et al., 2002). A relation between the shear deformation and the principal tensile stress in
the panel region was found and transformed into a moment-rotation relation to be assigned to the
rotational spring. The shear deformation is assumed to be equal to the rotation of the spring and the
moment is deduced as corresponding to the principal tensile stress evaluated on the basis of Mohr
theory. On the basis of the experimental results a first cracking level for external joints can be found

corresponding to a principal tensile stress value of pt = 0.2 f c ' . For internal joint the same

phenomenon occurs at a value of pt = 0.29 f c ' . While exterior joints show a strength degradation in

the post-cracking behavior, interior joints are characterized by a hardening behavior up to a value of

principal tensile stress of pt = 0.42 f c ' .

- 24 -
Chapter 4 – Modeling issues

Following these considerations the moment-rotation relationships governing the monotonic


behavior of the joint rotational spring model have been defined:

• Exterior joints: cracking occurs at a moment value corresponding to a pt = 0.2 f c ' .

The post-cracking behavior is described by a elasto-plastic relation without taking into


account the strength degradation.

• Interior joints: cracking occurs at a moment value corresponding to a pt = 0.29 f c ' .

The ultimate limit state corresponds to a value of pt = 0.42 f c ' . In this case a bi-linear

relationship is adopted to represent the hardening behavior of interior joints.


In figure 4.12 the moment-rotation relationships previously described are shown.

pt pt
0.42√fc'

0.29√fc'

0.20√fc'

0.00015
γ 0.0007 0.00015
γ 0.0007

a) b)

Figure 4.12: pt-γ relationships proposed for exterior (a) and interior joints (b)

This type of model doesn't take into any account the change in the joint resistance due to the
variation of axial load when the structure is subjected to cyclic lateral loading. The rotational spring
in fact connects two nodes that are respectively the conjunction of beams and columns, in this way
the axial load passes through the column elements without affecting the spring element. To solve
the problem a modification of the previous model has been proposed, the spring has been split in
two elements that are interposed between the beams connection node and the upper and lower
column respectively.

- 25 -
Chapter 4 – Modeling issues

Figure 4.13: Modified joint model representation (Trowland, 2003)

In the model the upper column end is slaved to the lower column end in lateral translation and
rotation. The beam is connected to the column through the joint rotational spring. The beam is
essentially continuous through the joint. This arrangement means that the joint forms a link in the
chain of elements through which the loads must be transmitted. The column can not transmit any
load to the beam except through the joint and vice versa. This solution also provides an axial load
path which passes through both the column and the joint. It is essential that axial load be transmitted
through the joint because the joint behaviour is dependent on the level of axial load. The joint must
be split into two so that it can be both an axial load path and a link between the column and beam.
The two springs are identical and both have half of the joints strength and stiffness.

4.3. MODELLING OF MASONRY INFILLS

The presence of masonry infills can drastically change the global seismic response of reinforced
concrete frames. If on one hand the infill panels will increase the stiffness and strength of the
building, on the other hand some unavoidable irregularities in both geometrical and mechanical
distribution of infills can lead to undesirable failure mechanisms that can compromise the bearing
capacity of the structure.
The correct modeling of masonry infills becomes therefore a basic issue for a realistic evaluation of
the seismic response of existing buildings. Several models have been proposed in literature to model
un-reinforced masonry infills. A rough classification can be obtained distinguishing between the
level of complexity of the model and the ability of capturing alternative failure mechanisms in the
infill panel together with local effects caused by the interaction with the frame.

- 26 -
Chapter 4 – Modeling issues

As for structural members and for joints, finite element models (F.E.M.) are available to model
masonry infills too. This approach can provide more complete results which can capture both the
global response and the local effects between the panel and the surrounding frame elements but
becomes less suitable when a large number of analyses have to be performed because of the high
computational effort requested. Another aspect that makes not worth using a highly refined model
for masonry infills is the elevated level of scatter of mechanical properties of masonry, especially in
existing buildings, that increases uncertainties in the definition of the model.
For these reasons an approach based on the use of axial springs acting as equivalent compression
diagonal struts has been widely adopted. In the following paragraph an overview of the available
multi-spring models is presented focusing the attention on the solution adopted in the present work
and on the method used to describe the mechanical properties of the equivalent struts.

4.3.1 Multi-Spring Models

Although it has been frequently modified, the diagonal strut model is the solution most widely
adopted to represent the interaction between masonry infills and reinforced concrete frame, because
of its simplicity and reliability. The simplest option is the use of two diagonal compressive struts
connecting centre to centre of the panel zone (Figure 4.14).

Compressive
Biella Strut
compressa

Figure 4.14: Equivalent diagonal strut model

The biggest limit of this modeling solution is that it doesn't allow to describe accurately the local
effects of the frame-panel interaction. For this reason some variations to the classical model has
been proposed, as shown in Figure 4.15.

- 27 -
Chapter 4 – Modeling issues

Model A
Modello

Aw

Modello
Model BB Modello
Model C
hz = z / 3

hz = z / 2
Aw / 4
Aw / 2

Aw / 2

Figure 4.15: Possible variations of the classical equivalent diagonal strut model

The two alternative models proposed (B and C) try to describe the effective stress migration along
the contact length between the panel and the structural elements when the building experiences a
lateral deformation. As shown by Crisafulli (1997), comparing the results obtained using the models
described above with those obtained from a finite element analysis it was found that the model C is
best describing the effective distributions of moments and shears in the structural elements, even if
unavoidable differences are still present caused by the concentrated application of the strut reaction
to columns and beams.
The models described up to now are not able to describe adequately the response of infilled frames
governed by shear failure with horizontal sliding. For this purpose Leuchars and Scrivener (1973)
propose a model made of two struts, which transfer the bending moment at mid-height of the
columns, and of a spring linked to the struts, that allows to describe the effect of friction developing
along the crack (Figure 4.16).

- 28 -
Chapter 4 – Modeling issues

Friction
Attrito

Figure 4.16: Sliding shear infill model (Leuchars and Scrivener, 1973)

4.3.2 Mechanical Properties of the Diagonal Strut

Since the equivalent diagonal strut model is widely adopted, several studies has been done with the
objective of defining some empirical relationships to evaluate the parameters which govern the
monotonic and cyclic behavior of the diagonal strut, as a function of the mechanical and
geometrical characteristics of the masonry. The basic parameters to be defined are the stiffness and
strength of the panel, for a non-linear monotonic analysis and of course the hysteretic behavior if a
non linear dynamic analysis has to be performed.

Stiffness

Classically the diagonal strut is connected to the intersection points of the geometrical axes of
structural elements constituting the frame. This means that the strut length is slightly bigger then the
effective length of the infill panel diagonal. It is generally recognized though that this discrepancy is
not affecting the reliability of the seismic response evaluation.
Usually the thickness and the elastic modulus of the strut are taken equal to those typical of the
masonry (although the Young Modulus can be modified taking into account the inclination of the
strut respect to the horizontal) , so that the only parameter to be defined to evaluate the axial
stiffness is the height of the cross section bw.
Holmes (1961) proposes a value of bw equal to one third of the diagonal length of the panel.
Stafford Smith (1969, 1996), on the basis of experimental tests, suggests values of the ratio bw/dw ,
where dw is the diagonal length of the panel, between 0.1 and 0.25. Paulay and Priestley (1992)
notice that an overestimation of bw can lead to excessive values of the global stiffness of the system

- 29 -
Chapter 4 – Modeling issues

resulting in unrealistically high seismic forces. For this reason they suggest to use bw equal to the
25% of dw.
The correct definition of the bw/dw ratio should take into account the correlation existing with the
infill-frame contact length and the masonry panel conditions. A fundamental aspect is cracking
level reached by the infill. When the panel is uncracked, in fact, the height of the strut cross section
can be relevant, but it can reduce to very limited values when the masonry panel is close to the
failure. Several studies based on the analogy of a beam on elastic support has been performed
starting from the sixties. Usually the beam corresponds to the frame column and the elastic support
to the infill panel.
One of the first researcher investigating this phenomenon was Stafford Smith (1969) who studied
the behavior of steel infilled frame with square shape and compared the experimental results with
the theoretical contact length evaluated as:
π
z= (4.2)

where λ is a factor which defines the relative stiffness between frame and panel:

Ewtw sin(2θ )
λ=4 (4.3)
4 Ec I p hw

with Ew and Ec elastic moduli of masonry and concrete, hw and tw height and thickness of the panel,
Ip moment of inertia of the columns cross section and θ the panel inclination respect to the
horizontal.
The height of the equivalent strut cross section is calculated as:
π
bw = 2 z sin(θ ) = sin(θ ) (4.4)
λ
Klingner and Bertero (1976, 1978), developing a previous relation (Mainstone, 1974), propose the
following formula to evaluate the bw/da ratio both for concrete and masonry infills:
bw
= 0.175(λ h) −0.4 (4.5)
dw
Bertoldi, Decanini and Gavarini (1993) propose a model applicable to masonry panels with and
without openings on the basis of results obtained from experimental tests on infilled frames and
from finite element non-linear analyses. The ratio bw/dw is calculated as representative of the
complete cracking level of the masonry panel subjected to cyclic loading and is evaluated through
the following expression:
bw K1
= + K2 (4.6)
dw λh

- 30 -
Chapter 4 – Modeling issues

where the parameters K1 and K2 are expressed as function of the product λh as shown in Table 4.1.

λh < 3.14 3.14 < λh < 7.85 λh > 7.85


K1 1.3 0.707 0.47
K2 -0.178 0.01 0.04
Table 4.1: Parameters of the equivalent diagonal strut model (Bertoldi, Decanini, Gavarini, 1993).

The previous equations require the evaluation of the elastic modulus of masonry Ew that is
necessary to define the axial stiffness of the equivalent strut. In an anisotropic material subjected to
a bi-axial tensile stress the elastic modulus in the direction inclined θ respect to the horizontal is
(Sacchi Mandriani et al., 1982):
cos 4 θ sen 4 θ ν
Eθ = [ ]−1
1
+ + cos 2 θ * sen 2 θ * ( − 2 ) (4.7)
E wh E wv G E wv
where Ewh , Ewv are the masonry elastic moduli in the horizontal and vertical direction respectively,
G is the masonry shear modulus, θ the inclination of the diagonal respect to the horizontal direction
and ν the Poisson coefficient.

Strength

The failure mechanisms of a masonry infill are multiple, for this reason the most correct approach is
to calculate a strength value associated to each mechanism and assume for the equivalent diagonal
strut the lowest of the values obtained, considering it as the resistance corresponding to the most
probable failure mechanism for the panel.
Bertoldi, Decanini and Gavarini (1993) identify four different possible failures: compression at the
center of the panel, compression of corners, sliding shear failure and diagonal tension. To each of
these phenomena a value of ultimate stress σw is associated and considered constant on the cross
section of the strut. The horizontal projection of the ultimate load corresponding to each mechanism
is calculated as:
Fw = σ wtwbw cos θ (4.8)

The equivalent strength σw for the four mechanism considered are evaluated with the following
equations:

- 31 -
Chapter 4 – Modeling issues

1.16 f w' tan θ


σw = compression at center of panel
K1 + K 2 λ h
1.12 f w' sin θ cos θ
σw = compression of corners
K1 (λ h) −0.12 + K 2 (λ h)0.88
(1.2sin θ + 0.45cos θ ) f wu + 0.3σ v
σw = sliding shear (4.9)
bw
dw
0.6 f ws + 0.3σ v
σw = diagonal tension
bw
dw

where fwu is the sliding resistance of the mortar joints and σv is the vertical compression stress due
to gravity loads, fws is the shear resistance under diagonal compression and f'w is the compression
resistance of the material.

Cyclic Behavior

In the present work the cyclic behaviour of the infill panel has been modelled adopting the
hysteretic rule proposed by Crisafulli (1997).to simulate the axial response of masonry. This model
takes into account the non-linear response of masonry in compression, including contact effects in
the cracked material (pinching) and small cycle hysteresis. This model also allows to take into
account the variation of strut's cross section as function of the axial deformation experienced by the
element. In this way is possible to consider the loss of stiffness due to the shortening of the contact
length between frame and panel as the lateral load increases. In Figure 4.17 the stress-strain
relationship and the backbone curve for the hysteretic model proposed are shown.

a) b)

Figure 4.17: Histeretic cycle (a) and backbone curve (b) of Crisafulli model (1997)

- 32 -
Chapter 5 – Validation of the analytical model

Equation Section (Next)

5. VALIDATION OF THE ANALYTICAL MODEL

The model of the joint shear hinge has been first validated through a series of analytical –
experimental comparisons with the results obtained from tests on beam-column subassemblies
performed at the Department of Structural Mechanics of the University of Pavia (Pampanin, 2002).
The same experimental program was completed with a test on a 2/3 scaled three storey reinforced
concrete bare frame, the results of this test has been compared with the analytical results obtained
from numerical analyses of a frame model where beam-column joints were modeled with the
proposed joint rotational spring model.
In a second phase the accuracy of the equivalent diagonal strut model proposed to represent the
masonry infill panels has been evaluated through comparison with the experimental pseudo-
dynamic tests on a series of 1/2 scaled one-storey one-bay frames performed at the Structural
Laboratory of the University of L'Aquila (Colangelo, 1999; 2003). Finally efficiency of the infills
model was checked comparing the analytical results with the results of experimental pseudo-
dynamic tests on a full scale four storey three dimensional R.C. frame carried out at the European
Laboratory for Structural Assessment (ELSA) of the Joint Research Centre of Ispra (Negro et al.,
1995).
In this chapter a more detailed description of the analytical – experimental comparisons used to
validate the reliability of the lumped plasticity model adopted in the present work is presented.

- 33 -
Chapter 5 – Validation of the analytical model

5.1. BEAM-COLUMN SUBASSEMBLIES

A series of different typologies of 2/3 scaled one-way beam columns subassemblies specimens were
tested at the Department of Structural Mechanics of the University of Pavia. They were chosen to be
representative of the different typologies of connections existing in a reinforced concrete frame
structure such as:

• Knee joints (specimen L1 and L2)

• Exterior tee-joints (specimen T1 and T2)

• Interior joints (specimen C1 to C4)


Within each joint type the beam longitudinal reinforcement was varied between the specimens and
for the interior joints different anchorage details were assumed for the longitudinal rebars passing
through the joint region.

5.1.1 Material properties

Since the specimens were intended to be representative of older structures, designed and built
between 1950 and 1970, the materials used were chosen in accordance with the design practice of
the time. A low-strength concrete Rck 200 was specified (which refers to the Italian standard and
corresponds to a cube compression strength of around 20 MPa). However significantly higher
values of the concrete average compression strengths at 28 days were obtained. Steel smooth bars,
with mechanical properties (allowable stress 160 MPa) similar to those typically used in that period,
were adopted for both longitudinal and transverse reinforcements. Table 5.1 shows the average
mechanical properties of materials used in the tests.

CONCRETE STEEL (longitudinal bars)


Cylindric compression Cube compression Yielding Ultimate
strength (MPa) strength (MPa) (MPa) (MPa)
Average Average Diameter φ 8
23.9 (0.52) 29.1 (0.64) 385.6 (1.75) 451.2 (3.49)
Diameter φ 12
345.9 (2.17) 458.6 (2.17)

Table 5.1: Specimen material properties (Pampanin, 2002)

- 34 -
Chapter 5 – Validation of the analytical model

5.1.2 Knee joint

The knee joint taken in consideration is the one named L1. The specimen has been built trying to
reproduce the real boundary conditions of joint in a reinforced concrete frame building. Beam and
column elements were extended between contraflexure points (assumed to be at midspan of the
beams and at midheight of the columns) and connected with pins to the ground. To simulate the
simple support at the beam end a pin-end steel member was placed to connect the beam to the floor.
The beams and columns scaled dimensions were 330mm (depth) x200mm (width) and 200x200mm,
respectively. In figure 5.1 the geometric dimensions and the reinforcement layout of the L1
specimen are shown.

Figure 5.1: Specimen L1: geometric characteristics and steel bars layout (Pampanin, 2002)

Table 5.2 reports a summary of the reinforcement present in the beam and column.

BEAM
SPECIMEN COLUMN REINFORCEMENT
REINFORCEMENT
Top 2Φ8+2Φ12
L1 3 Φ8 + 3 Φ8
Bottom 2Φ8+2Φ12

Table 5.2: Specimen L1 reinforcement (Pampanin, 2002)

- 35 -
Chapter 5 – Validation of the analytical model

Cyclic horizontal loading was applied to the beam end using a hydraulic actuator in displacement
control (figure 5.2).

Figure 5.2: Test setup of specimen L1 (Pampanin, 2002)

- 36 -
Chapter 5 – Validation of the analytical model

5.1.3 Interior joint

The specimen considered as representative of the interior joint typology (cruciform joint) is the C2.
The geometrical dimensions of beams and columns are the same as those used for the knee joint
test. In this specimen the beam reinforcement is running continuously through the panel joint
region. The same boundaries conditions as for the specimen L1 were adopted to simulate the
presence of a surrounding frame. Furthermore, in addition to the horizontal load applied to the top
of the upper column, a vertical load has been applied to describe the axial load variation in columns
happening during a seismic event. The variation law of the vertical load was taken as a linear
function of the lateral load applied, as described by the following relation:
N = −120 + 1.4 ⋅ Fh [kN ] (5.1)
The axial load was applied by means of a vertical hydraulic jack, acting on a steel plate connected
to the column base plate by vertical external post-tensioned bars.
The geometrical dimensions and the reinforcement layout for the specimen C2 are shown in figure
5.3 and table 5.3.

Figure 5.3: Specimen C2: geometric characteristics and steel bars layout (Pampanin, 2002)

BEAM
SPECIMEN COLUMN REINFORCEMENT
REINFORCEMENT
Top 2Φ8+2Φ12
C2 3 Φ8 + 3 Φ8
Bottom 2Φ8+1Φ12

Table 5.3: Specimen C2 reinforcement (Pampanin, 2002)

- 37 -
Chapter 5 – Validation of the analytical model

Figure 5.4 shows the test setup for the interior joint typology (specimen C2).

Figure 5.4: Test setup of specimen C2 (Pampanin, 2002)

- 38 -
Chapter 5 – Validation of the analytical model

5.1.4 Exterior joint

Both exterior joint specimens tested (T1 and T2) have been considered in the numerical-
experimental comparison. The geometric dimensions of beam and columns are the same as the
previous specimens (knee and interior joints) and are shown in figure 5.5.

Figure 5.5: Specimen T1: geometric characteristics and steel bars layout (Pampanin, 2002)

The two specimens differ only in the beam longitudinal reinforcing steel ratio. This choice was
made to investigate the different behavior of the joint in the case the beam is yielding (T2) or not
(T1). In table 5.4 and figure 5.6 the reinforcing steel amount and disposition is reported.

COLUMN BEAM
SPECIMEN
REINFORCEMENT REINFORCEMENT
Top 2Φ8+2Φ12
T1 3 Φ8 + 3 Φ8
Bottom 2Φ8+2Φ12
Top 2Φ8+1Φ12
T2 3 Φ8 + 3 Φ8
Bottom 2Φ8+1Φ12

Table 5.4: Specimen T1 and T2 reinforcement (Pampanin, 2002)

- 39 -
Chapter 5 – Validation of the analytical model

Column Section
Sezione pilastro Beam Section
Sezione trave(T1) Beam Section
Sezione trave(T2)

Figure 5.6: Reinforcement layout of specimens T1 and T2 (Pampanin, 2002)

As for the case of interior joint a vertical load variation has been imposed to the top of the upper
column through an hydraulic jack. The axial load (N) varies linearly with the lateral force (Fh)
applied to the specimen, starting from a value of 100 kN taken as representative of the gravity load
effect, following the relationship:
N = −100 + 2.44 ⋅ Fh [kN ] (5.2)

The test setup for the exterior joint specimens T1 and T2 is shown in the figure 5.7.

Figure 5.7: Test setup of specimen T1 and T2 (Pampanin, 2002)

- 40 -
Chapter 5 – Validation of the analytical model

5.2. SPECIMENS MODELLING

5.2.1 Structural elements

Beams and columns has been modeled with two dimensional linear elements. The inelastic behavior
is concentrated at the member ends in a plastic hinge region with a length defined by the following
formula (Priestley et al., 1996):
L
L p = 0.08 + 0.022 f yl dbl (5.3)
2
The monotonic behavior has been defined starting from a moment-curvature analysis of the section.
The elastic stiffness is taken as the secant stiffness at yielding point. To describe the cyclic behavior
of the section the Takeda (Takeda, 1970; Otani, 1974) hysteresis rule has been used (see figure 4.3
a). An axial load-bending moment interaction diagram has been assigned to the column member to
take into account the effect of axial load variation, due to cyclic lateral loading, on the column
strength.
The column is connected to the ground with a hinge and the beam is simply supported with free
horizontal translation in order to represent the real restraint conditions of the specimens tested.

5.2.2 Joint elements

The beam-column joint has been modeled with a couple of rotational spring as described in
paragraph 4.2.2.

Figure 5.8: Modified joint model representation (Trowland, 2003)

- 41 -
Chapter 5 – Validation of the analytical model

Figure 5.8 schematically shows the model adopted. The springs are actually zero length elements
and the effective dimensions of the panel zone region are modeled with rigid and blocks. The upper
column end is slaved to the lower column end in lateral translation and rotation. The spring
elements used are identical and they both have half of the joint strength and stiffness. The elastic
rotational stiffness of the joint spring is calculated as:
⎛ 0.9d b H ⎞
K = G⎜ ⎟ Ac (5.4)
⎝ H − 0.9d b ⎠
with:
G: concrete shear modulus
Ac: column cross section area
H: interstorey height
db: beam depth
The springs have the same axial stiffness of the columns connecting to the joint and is calculated as:
EAc
K= (5.5)
L
with:
E: concrete elastic modulus
Ac: column cross section area
L: half of the joint panel height
As described in paragraph 4.2.2 the cracking moment of the joint has been calculated as

correspondent to a principal tensile stress value of 0.2 f c ' for exterior and knee joints and 0.29 f c '

for interior joint. After the cracking point interior joints show an hardening behavior up to a level of

pt = 0.42 f c ' .

5.2.3 Joint hystetesis rule

The cyclic behavior of the joint rotational spring has been defined using an hysteretic rule available
in the RUAUMOKO program and specifically proposed to describe the characteristics of the joint
response. In particular the adopted hysteretic loop is able to describe the typical "pinching" effect
due to the slippage of plain round reinforcing bars through the joint panel zone and to the opening
and closing of diagonal shear cracks in the joint region. The hysteresis rule needs the definition of
six parameters governing the unloading and reloading phases of the cycle and in addition is possible
to chose between two different options for the definition of the reloading branch (figure 5.9).

- 42 -
Chapter 5 – Validation of the analytical model

dp
F
βdp
Option 1
∆F
rK0 Ku1=K0
Ks2 µαu1

K0
Ku2=K0
Ks1 µαu2 d

Ku2
Ks1= K0
Ks2
µαs1
Ks2= K0
∆F Ku1 rK0
µαs2

a)

dp
F
Option 2 βdp

∆F
rK0 Ku1=K0
µαu1
K0
Xidr
Ku2=K0
dr Ks1 µαu2 d

Ku2
Ks1= K0
µαs1
Ks2= K0
Ku1 rK0
µαs2

b)

Figure 5.9: Pampanin histeretic rule (Carr, 2004): a) Option 1-reloading power factor; b) Option 2-reloading slip
factor

The monotonic branch is a bi-linear curve where the initial stiffness K0 is the cracked stiffness of
the joint and the second branch represent the hardening phenomenon noticed during the tests on
interior joints. The unloading path is described by two lines, the first is characterized by a stiffness

- 43 -
Chapter 5 – Validation of the analytical model

K0
Ku1 = where µ is the ductility level reached in the last cycle and αu1 is the initial unloading
µα u1

power factor and it develops along a force interval defined by the unloading force factor ∆F, defined
as percentage of the yielding force. The second unloading branch is governed by the stiffness
K0
Ku2 = where the final unloading power factor αu2 is another parameter of the cycle. The
µα u2

reloading phase of the hysteresis rule is still defined by two braches with different stiffness.
Considering that the reloading target point is known and defined by the reloading factor β as a
percentage of the plastic deformation reached in the previous cycle, the reloading process can be
completely defined by two parameters. This hysteresis rule allows the choice between two reloading
options. The option 1 (figure 5.9a) needs the definition of two stiffness, the first (Ks1) governing the
slipping reloading branch and the second (Ks2) which defines the second reloading phase up to the
target point βdp. The stiffness Ks1 and Ks2 are calculated as:
K0 K0
Ks1 = αs 1
Ks 2 = (5.6)
µ µα s2

where m is the ductility level reached in the last cycle and αs1 and αs2 are empirical parameters.
In the option 2 the reloading phase is described by the stiffness of the first branch Ks1 and by the
slippage length, defined as follows:
K0
Ks1 = δ s = Xidr (5.7)
µα s1

where Xi is the reloading slip factor and dr is the residual displacement obtained after the last
unloading phase. It is so assumed that the slippage length is directly proportional to the residual
displacement that is an index of the damage level reached by the joint. In this way the more
extensive is the damage of the joint (i.e. the wider are the cracks in the joint panel region) the more
evident is the "pinching" behavior in the hysteresis loop.
In table 5.5 the parameters necessary to define the hysteretic rule are listed for both versions,
together with an indication of the process to which they are associated.

Option 1 Option 2 Related Process


αs1 αs1 reload
αs2 Xi reload
αu1 αu1 unload
αu2 αu2 unload
∆F ∆F unload
β β reload
Table 5.5: Parameters needed to define hysteresis rule adopted for joint members

- 44 -
Chapter 5 – Validation of the analytical model

The calibration of the parameters of the cycle has been made on the basis of the results of the
experimental tests on the beam-column subassemblies for both the versions of the hysteresis rule. In
the present work it was chosen to use the second option in which the relation between the damage
and the definition of the slippage length is more direct to recognize.
Table 5.6 summarizes the values of the parameters obtained from the calibration process and used
in the numerical analyses performed in the present work.

Specimen
Parameter
T1 T2 L1 C2
αs1 1.2 1.3 1.3 1.2
Xi 1.5 1.3 1.4 1.3
αu1 -0.1 -0.1 -0.1 -0.1
αu2 0.9 0.8 0.8 0.95
∆F 30 30 20 30
β -0.2 -0.3 -0.1 0
Table 5.6: Calibration of the hysteretic rule parameters for the beam-column subassemblies

The RUAUMOKO program (Carr, 2004) allows the definition of a strength degradation relationship
that can be either function of the ductility or the number of load reversal from the backbone or spine
curve of the hysteresis rule. In particular a ductility based strength degradation has been assigned to
the exterior joint model to describe the loss of resistance showed by the T1 and T2 specimens
during the experimental tests. No strength degradation is used for the knee and interior joint
typologies.

5.2.4 Limit states for joints

As observed in the experimental tests and confirmed by the numerical investigations, the occurrence
of damage in the joint region can result, (through activation of a shear hinge) to a reduction of the
interstorey drift demand, thus postponing or avoiding the development of a soft storey mechanism.
On the other hand, the increased shear deformation demand in the joint region can lead, depending
on the joint type and on the structural details adopted, to a sudden strength reduction with loss of
vertical-load bearing capacity. The maximum joint shear deformation γ has thus to be considered a
primary parameter to be monitored during numerical analysis and compared with reference values
corresponding to different limit states in order to appropriately define the structural performance or
damage level. By comparing the damage observed in the experimental tests with the level of joint

- 45 -
Chapter 5 – Validation of the analytical model

rotation obtained in the spring model, when reproducing the overall hysterisis behaviour,
preliminary values corresponding to Limit States (LS) or Performance Level (PL) related to the
damage in exterior joints can be tentatively suggested as follows (Pampanin et al., 2002):
PL1: first cracking and limited damage 0.0002 < γ < 0.005
PL2: extensive damage 0.005 < γ < 0.01
PL3: critical damage (reparability issues arise) 0.01 < γ < 0.015
PL4: incipient collapse γ > 0-015

5.2.5 Analytical – experimental comparison

A series of three cycles at increasing level of interstorey drift was applied thought the horizontal
hydraulic actuator. The general loading time history is illustrated in figure 5.10.
Since the model’s strength degradation was based on ductility and not inelastic cycles, only one
cycle was made at each level of drift in the numerical simulation.
4
3.5%
3%
3
2%
Imposed Top Drift (%)

2 1.5 %
0.8 % 1%
1 0.6 %
0.2 %

-1

-2

-3

-4
10 20 30 40 50
n. of semi-cycles

Figure 5.10: Typical loading history (Pampanin, 2002)

Knee joint specimen

In the knee-joint specimen L1 the damage is mainly concentrated at the column interfaces with
crushing and spalling of the concrete at the top face of the joint zone at higher level of drift, due to
slippage of the column reinforcement and stress concentration at the end-hook.

- 46 -
Chapter 5 – Validation of the analytical model

Top Displacement[mm]
-60 -40 -20 0 20 40 60
15
L1 JOINT
10

Lateral Force [kN]


5

-5
Experimental
-10 Analytical

-15
-4 -3 -2 -1 0 1 2 3 4
Top Drif t [%]

Figure 5.11: Analytical-experimental comparison for specimen L1

The comparison between the experimental results (grey line) and the numerical values obtained
from the analysis (black line) show a good agreement both in the description of the unloading phase
and the "pinching" phenomenon characterizing the reloading branch and due in this case to the
slipping of the column reinforcing bars. Furthermore the axial load-bending moment interaction
associated to the joint spring allows to describe the difference in the value of the cracking moment
in the positive and negative directions.

Interior joint specimen

The interior joint specimen C2, taken into consideration for the scope of this work, is characterized
by continues longitudinal reinforcing bars through the joint region. The behavior of the beam-
column subassemblies is governed by the yielding of the column. After an initial cracking at both
column-to-joint and beam-to-joint interfaces, flexural damage concentrated at the column end,
widening the existing interface crack and leading (from 2.0 % drift level) to progressive damage.

- 47 -
Chapter 5 – Validation of the analytical model

Top Displacement[mm]
-80 -60 -40 -20 0 20 40 60 80
20
15 C2 JOINT
10

Lateral Force [kN]


5
0
-5
-10 Experimental
Analytical
-15
-20
-4 -3 -2 -1 0 1 2 3 4
Top Drif t [%]

Figure 5.12: Analytical-experimental comparison for specimen C2

Also in this case the adopted hysteretic loop leads to a good agreement in results for all values of
interstorey drift.

Exterior joint specimen

The specimen T1 is characterized by an amount of longitudinal reinforcement ratio in the beam high
enough to make the joint the weak element of the subassembly. For this reason the response of the
specimen is governed by the joint that is the only damaged element. The fundamental source of the
peculiar damage mechanism was related to the concentration of compression force of beam
reinforcing bars at the hook-end, after premature loss of bond strength within the joint region.
The T2 specimen shows the formation of a plastic hinge in the beam (with a lower longitudinal
reinforcement ratio respect to the specimen T1) and the subsequent formation of a "shear hinge" in
the joint. The cyclic response of both specimens is characterized by a strength degradation
phenomenon that is more evident for the specimen T2.

- 48 -
Chapter 5 – Validation of the analytical model

Top Displacement[mm] Top Displacement[mm]


-80 -60 -40 -20 0 20 40 60 80 -80 -60 -40 -20 0 20 40 60 80
15 15

T1 JOINT 10 T2 JOINT
10
Lateral Force [kN]

5 5

0 0

-5 -5
Experimental Experimental
Analytical
-10 Analytical -10

-15 -15
-4 -3 -2 -1 0 1 2 3 4 -3 -2 -1 0 1 2 3
Top Drift [%] Top Drif t [%]

Figure 5.13: Analytical-experimental comparison for specimen T1 and T2

The model adopted describes well the "pinching" effect typical of the specimens response and due
to the opening of diagonal cracks in the joint panel region and to the slipping of longitudinal steel
bars in beams and columns connecting to the node. The capability of the model to take into account
the effect of the variation of axial load on the strength of the spring element allows to describe the
difference between the positive and negative value of the cracking force. In the modeling of the
exterior joint specimens a strength degradation relationship, function of the ductility, has been used
calibrated on the effective loss of strength of joints recorded during the tests.
It has to be noticed that, because of the bi-linear idealization of the moment rotation relationship of
the joint, there is some discrepancy between the initial uncracked stiffness of the specimens and the
elastic stiffness of the models. It is believed that the lower level of accuracy in the description of the
elastic behavior of the joints can be acceptable when a non-linear dynamic analysis of a frame
structure is performed, since the cracking of the joint panel zone is likely to occur early during a
seismic event.

- 49 -
Chapter 5 – Validation of the analytical model

5.3. THREE STOREY FRAME

The joint model calibrated on the beam-column subassemblies was then used to represent the effect
of the cyclic behavior of joints on the global seismic response of a reinforced concrete frame
building. The validation of the model was made comparing the numerical results with the results
obtained from an experimental test performed at the Laboratory of the Department of Structural
Mechanics of the University of Pavia.
As part of a coordinated national project on the seismic vulnerability of existing reinforced concrete
frame buildings designed for gravity loads only, as typical in Italy between the 1950's and the
1970's, a quasi-static cyclic experimental test was performed on a 2/3 scaled three storey frame.

5.3.1 Frame geometry and reinforcement details

The prototype building was a multiple frame system consisting of a three-story (3 m height) and
three-bays (4.5, 2 and 3.5 m respectively). The column sections were 300x300 mm and the beam
sections 500x300mm (depth x width).
The total height of the 2/3 scaled test frame was 6 m with interstorey height of 2 m. The bays were
2.81 m, 1.14 m and 2.13 m long respectively. The scaled section dimensions of columns and beams
were 200x200mm and 330mm (depth) x 200mm (width), respectively. The design
recommendations provided by the current national design provisions, integrated by text-books
broadly adopted in the engineering practice and available between the 1950’s and 1970’s were
followed. Consistently with the old practice, no transverse reinforcement were placed in the joint
region. Plain round bars, with mechanical properties similar to those typically used in older periods,
were adopted for both longitudinal and transverse reinforcement. Beam bars in exterior joints were
not bent in the joint region, but anchored with end-hooks. Lap splices with hook anchorages were
adopted in the beam bars crossing interior joints as well as in column longitudinal bars at each floor
level above the joint region and at the column-to-foundation connection. The geometrical and
reinforcement characteristics, together with a detail of columns lap slices are reported in figures
5.14 and 5.15.

- 50 -
Chapter 5 – Validation of the analytical model

B1 B3 B3 B5 B5 B5

Column section C Beam section B1 Beam section B3 Beam section B5


20 mm 20 mm 2+2 Ø 8 20 mm 20 mm 2+2 Ø 8
3+3 Ø 8 2+2 Ø 8
200 mm

tie Ø 4
330 mm

330 mm
330 mm

tie Ø 4 tie Ø 4 tie Ø 4

200 mm
2+2 Ø 12 3 Ø 12 2+1 Ø 12
200 mm 200 mm 200 mm

Figure 5.14: Geometrical properties and reinforcement layout of the test-frame (Pampanin, 2003)

Figure 5.15: Column lap splice at floor level (units in cm) (Pampanin, 2003)

- 51 -
Chapter 5 – Validation of the analytical model

5.3.2 Material properties

Material used in the construction of the test frame have characteristics similar to those most widely
used in the period 1950-70.
Mean values and standard deviations for concrete cylinder compression strength, f’c, are indicated
in table 5.7.
f’c(MPa)
Concrete strength f’c (MPa) @28 days
@starting test-day
1st Floor Column 15.97 17.83
Beam 13.19 13.28
2nd Floor Column 12.63 13.19
Beam 12.82 13.84
3rd Floor Column 13.00 13.47
Beam 12.63 12.72
Mean Value 13.37 14.06
Standard Deviation 1.29 1.89
Mean(1st floor excluded) 12.85 13.30
Stand. Dev. (1st floor excl) 0.24 0.41
Table 5.7: Concrete cylinder compressive strength values (Pampanin, 2002)

It should be noted that, as typical of construction practice, the columns and beam elements at each
floor were casted at subsequent stages, leading to a high standard deviation in the mechanical
properties of the concrete. In particular the concrete strength values of the first storey columns
result higher than the upper floors members. After casting the first storey, the characteristic of the
concrete were in fact modified to increase the workability leading to the aforementioned lower
strength values for the next casting. Average values and standard deviations are thus shown in the
two cases: including and excluding the strength value of the first floor columns.
Plain round steel bars with characteristic yielding stress fy = 350 MPa were adopted for the
longitudinal and transversal reinforcement. From laboratory tests on steel bars the mean yielding
and ultimate strength values were established. Table 5.8 reports the main mechanical properties of
the longitudinal reinforcement.
Ultimate
Yielding
Bar Diameter (mm) Steel Strength f (MPa)
fy (MPa) u
8 Mean 385.64 451.22
Standard Deviation 1.74 3.48
12 Mean 345.87 458.63
Standard Deviation 2.17 2.17
Table 5.8: Mean steel strengths of longitudinal bars (Pampanin, 2002)

- 52 -
Chapter 5 – Validation of the analytical model

5.3.3 Test setup and loading history

The test frame was subjected to a quasi-static cyclic loading at increasing levels of top
displacement, applied to the structure using three actuators connected to the closest beam through a
steel level arm (figure 5.16).

Quasi-Static Gravity loads Screw Jack


Cyclic Actuators
Loadin

Reaction wall

Strong floor

Figure 5.16: Test setup (Pampanin, 2003)

- 53 -
Chapter 5 – Validation of the analytical model

The loading history consisted of a series of three cycles at increasing level of top drift ( ± 0.2%; ±
0.6%; ± 1.2%) with one conclusive cycle at ± 1.6%. The top floor displacement was directly
controlled while maintaining a constant ratio between the applied lateral forces accordingly to a
code type distribution proportional to the mass and the floor level height. Being the mass at the top
floor lower than those at the first and second floor, the following lateral force ratio was obtained:
⎧1 ⎫ ⎧1 ⎫ ⎧1 ⎫
{F } = ⎪F ⎪ ⎪ ⎪ ⎪ ⎪
⎨ i 2 Fi1 ⎬ = ⎨(m2 / m1 ) ⋅ h2 h1 ⎬ ≅ ⎨0.90⎬
i
(5.8)
Fi1 ⎪ F F ⎪ ⎪(m / m ) ⋅ h h ⎪ ⎪0.45⎪
⎩ i 3 i1 ⎭ ⎩ 3 1 3 1 ⎭ ⎩ ⎭

⎧ Fi1 ⎫
⎪ ⎪
{ }
where F i = ⎨ Fi 2 ⎬ is the vector of floor forces at the ith loading step, m j and h j (with j=1,3)
⎪F ⎪
⎩ i3 ⎭i
indicate the jth-floor mass and height (relative to the foundation level). Figure 5.17 shows the test
loading history and the lateral force distribution.
4
3.5%
3%
3
2%
Imposed Top Drift (%)

2 1.5 %
0.8 % 1%
1 0.6 %
0.2 %

-1

-2

-3

-4
10 20 30 40 50
n. of semi-cycles

Figure 5.17: Lateral loading history and force distribution (Pampanin, 2002)

The presence of gravity loads where simulated using concrete blocks supported by the beams and
arranged as shown in figure 5.16. Table 5.9 reports the values of gravity loads applied to the
structure for each bay and floor level.

1st Bay (2.8 m) 2nd Bay (1.13 m) 3rd Bay (2.13 m)


31.8 kN 14.1 kN 27.1 kN
1st floor
(4*570 kg+ 2*480 kg) (3*480 kg) (4*570 kg + 1*480 kg)
31.8 kN 14.1 kN 27.1 kN
2nd floor
(see first floor) (see first floor) (see first floor)
22.4 kN 9.4 kN 22.4 kN
3rd floor
(4*570 kg) (2*480 kg) (4*570 kg)
Table 5.9: Gravity load distribution (Pampanin, 2002)

- 54 -
Chapter 5 – Validation of the analytical model

5.3.4 Description of the model

The modeling approach used for the test frame follows the same principles of the one used to
represent the beam-column subassemblies and described in chapter 2. Figure 5.18 shows the
geometrical dimensions of the numerical model. The base nodes have been modeled as fully fixed
since the test frame foundation was constituted by a 300mm thick continuous beam anchored to the
laboratory ground floor by means of continuous steel girder beams, thus providing enough rigidity
to consider the structure fully fixed to the ground.

Figure 5.18: Geometric dimensions of the frame model

Beams and columns members are represented by mono-dimensional elements with inelasticity
concentrated in the end critical sections. The initial and post-yielding stiffness of beams and
columns elements were defined through a moment curvature analysis of each section. For column
members a series of moment-curvature analysis with increasing value of applied axial load were
performed in order to define a realistic axial load-bending moment interaction diagram to assign to
each element in the numerical model (table 5.10 and figure 5.19).

- 55 -
Chapter 5 – Validation of the analytical model

1st storey columns 2nd-3rd storey columns


N [kN] M [kNm] N [kN] M [kNm]
-110.1 0.0 -110.1 0.0
0.0 9.5 0.0 9.4
107.5 16.2 90.0 15.0
215.0 22.0 180.0 19.5
322.5 20.8 270.0 18.2
750.1 0.0 630.1 0.0
Table 5.10: Interaction diagram values for the frame columns

25
1st Storey
2nd-3rd Storey
20
Moment [kNm]

15

10

0
-200 0 200 400 600 800
Axial Load [kN]

Figure 5.19: Interaction diagrams of the frame columns

The non linear behavior of the beam members was described adopting a modified Takeda (Otani,
1974) hysteretic rule, setting the cycle parameters α = 0.5 (for the unloading phase) and β = 0 (for
the reloading branch). These values allows to minimize the area of the hysteretic loop thus limiting
the energy dissipated by the plastic hinges, as typically happens in under designed structural
elements.

- 56 -
Chapter 5 – Validation of the analytical model

Figure 5.20: Modified Takeda hysteresis rules(Otani,1974)

The columns members has been provided with Fukada (Fukada, 1969) tri-linear hysteresis to better
describe the effect of cracking on the element seismic response (figure 5.21). This solution has been
used for the pushover analyses performed on the frame. Performing the cyclic analysis, the presence
of a tri-linear loop led to a excessive increase of computational efforts, so it has been decided to
adopt a Takeda hysteresis both for beams and columns. This type of rule is numerically more stable
and since is supposed that the columns will crack for relatively small value of displacement (drift),
it has been assumed reasonable to use as elastic stiffness the secant stiffness at yielding point.

Figure 5.21: Fukada hysteresis rule (Fukada, 1969)

The beam-column joints are modeled with two axial-rotational spring interposed between the upper
and lower column as described in paragraph 5.2. The cracking moments for the joint panels have

been calculated as correspondent to a value of principal tensile stress pt = 0.2 f c ' and pt = 0.29 f c '

for exterior and interior joints respectively, as described in paragraph 4.2.2. The cyclic behavior of
the beam-column joints has been described using the hysteretic rule described in paragraph 5.2.3
- 57 -
Chapter 5 – Validation of the analytical model

whose parameters were calibrated on the basis of the results of the experimental test on beam-
column subassemblies as reported in the previous paragraphs.

5.3.5 Performed analyses

The frame model has been first studied performing two different pushover analyses (displacement
controlled and force-controlled) to investigate the monotonic response of the structure. A
displacement controlled cyclic analysis was then performed to compare the hysteretic response of
the frame with the experimental results obtained during the test and to investigate the cyclic
behavior of structural members and joints.

Displacement-controlled pushover

A vector of displacements with increasing amplitude has been applied to the exterior joints at each
floor level. The ratio between the floor displacement amplitudes was kept constant and equal to the
ratio between the experimental displacements recorded at 1.6% of top drift. Table 5.11 shows the
ratio between floor displacements applied to the model.
DOF ∆3/∆i
1 0.52
2 0.83
3 1
Table 5.11: Floor displacement ratio applied in the pushover analysis
where:
∆i is the displacement at the ith storey
∆3 is the displacement of the 3rd storey
The analysis was performed loading the structure both in the positive and negative directions.

Force-controlled pushover

The forces applied to each storey reflected the force ratio defined in the test setup and calculated as
function of the floor height and mass (see paragraph 5.3.3). The applied force ratio is reported in
table 5.12.
DOF F3/Fi
1 0.45
2 0.9
3 1
Table 5.12: Ratio of storey forces applied in the pushover analysis

- 58 -
Chapter 5 – Validation of the analytical model

where:
Fi is the force applied at the ith storey
F3 is the force applied of the 3rd storey
Also the force controlled pushover analysis was performed applying the loads first in the positive
and than in the negative direction.

Cyclic analysis

In the cyclic analysis the displacement history applied to the joints of the frame was the same used
in the experimental test: 3 cycles at increasing level of top drift (0.2%, 0.6% and 1.2%) and a final
cycle at 1.6% of top drift (see figure 5.22).

2 1.6%
Imposed Top Drif t [%]

1.2%
1 0.6%
0.2%
0

-1

-2

0 10 20 30 40
Test Sequence

Figure 5.22: Imposed displacement history for the cyclic analysis

5.3.6 Analytical-experimental comparison

Pushover Analysis

The pushover analysis was intended to investigate the monotonic response of the frame and
compare it with the envelope curve of the cyclic response of the test frame obtained from the quasi-
static experimental test. The numerical results have been then compared with a series of base shear
– top drift curves presented by the participants to a blind analytical prediction contest organized as
part of the coordinated research program. The contest consisted on the prediction of the expected
response of the frame system under simulated seismic loading and has been carried out within some
partners of the project (Third University of Rome, University on Naples) as well as some external
research teams either from academic (University of Canterbury, Christchurch, New Zealand) or
from consulting-industry (Rutherford & Checkene Engineers, San Francisco, USA).

- 59 -
Chapter 5 – Validation of the analytical model

Figure 5.23 shows the comparison between the analytical results of the displacement controlled
pushover, the blind prediction results and the experimental cyclic response of the frame.
Top Floor Drift (%)
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
80

60

40
Base Shear (kN)

20

Experimental
-20
Prediction- Naple
Prediction- Rome3
-40 Prediction- Canterbury
Prediction-Salerno
Prediction-Ruth.&Check.Eng.
-60 Analytical

-80
-120 -80 -40 0 40 80 120
Top Floor Displacement (mm)

Figure 5.23: Comparison between pushover curve (displacement control),


blind predictions and experimental results (after Pampanin, 2002)

The analytical results show a good agreement with the experimental response, in particular the
adoption of a tri-linear hysteresis rule to the columns members allows to describe well the initial
stiffness of the frame. The maximum base shear capacity was well estimated both in the positive
and in the negative direction. Any strength degradation rule was adopted for the structural elements
of the model (except that for the rotational springs of the exterior joints) and this is reflected by the
loss of accuracy in the description of the response at higher level of ductility. In is worth noting that
while the curves obtained by the competitors of the contest are blind predictions calculated knowing
the information on geometrical and mechanical characteristics, the numerical analysis has been
performed after the experimental test with the complete access to all experimental results.
In figure 5.24 the base shear – top drift curve calculated from the force-controlled pushover analysis
is presented. It can be noticed a slight overestimation of the maximum base shear capacity in the
positive direction while in the negative the value of the base shear is similar to that obtained from
the displacement controlled pushover. The displacement profile at increasing value of top drift is
also shown in the figure. The overestimation of the displacement is marked at higher level of top
drift (1.2% and 1.6%) and is due to the higher inelastic deformation demand of the joint springs
respect to the effective shear deformation suffered by the beam column joints during the test.

- 60 -
Chapter 5 – Validation of the analytical model

Top Floor Drif t (%)


-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
80

60

40
Base Shear (kN)

20

-20

-40

-60

-80
-120 -80 -40 0 40 80 120
Top Floor Displacement (mm)

Experimental
Analytical

Top floor drift (%)


-1.8 -1.2 -0.6 0 0.6 1.2 1.8

3 6
Height (m)
Storey

2 4

1 2

0 0
-120 -80 -40 0 40 80 120
Displacement profile (mm)

Analytical Profile
Experimental Profile

Figure 5.24: Force-controlled pushover: base shear – top drift curve and displacement profile (after Pampanin,
2002)

- 61 -
Chapter 5 – Validation of the analytical model

The experimental test on the frame system confirmed the high vulnerability of the panel zone region
and the tendency to develop undesirable global mechanism due to the absence of an adequate
hierarchy of strength. Figure 5.25 show an overview of the structural damage pattern correspondent
to a top drift level of 1.2%.
Top Displacement

Shear Hinges

Plastic Hinges

Figure 5.25: Crack pattern observed at 1.2% top drift (Pampanin, 2003)

- 62 -
Chapter 5 – Validation of the analytical model

Figure 5.26: Frame damage photo report (Pampanin, 2002)

It can be noted that damage mostly concentrated in the joint region (exterior tee-joints) or at the
beam/column interfaces with a wide flexural crack as expected due to plain round bars slip. In the
interior joint panel regions no cracks were observed. The damage pattern that results from the
displacement controlled pushover analysis is presented in figure 5.27.

- 63 -
Chapter 5 – Validation of the analytical model

Loading Top Drift 1.2% Top Drift 1.2% Loading


Direction Direction

1.88 2.57 2.55 2.25 2.66 3.74 3.78 2.94

1.24 1.96 1.89 1.22 1.55 4.13 4.61 3.40

1.57 2.53 2.57 1.91 1.2 3.59 3.92 2.88

2.02 2.71 2.81 2.70 3.29 4.25 3.95 3.58

Figure 5.27: Damage pattern and curvature ductility at 1.2% top drift level (displacement controlled pushover)

The plastic hinge distribution on the numerical model frame reflects the effective damage pattern
observed during the experimental test. It is worth noting in particular that the exterior joint spring of
the first and second storey experienced the higher damage while the interior joint didn't reach the
cracking limit. Furthermore the higher curvature ductility demands are recorded at the base column
sections and at the upper section of the 2 storey columns. At the first storey level, in fact , the
inelastic deformation was concentrated into the joint springs, thus preventing higher plastic
deformations in the column sections. Similar results were obtained from the force controlled
pushover (figure 5.28), in this case a general increase of ductility demand in the column members is
noticed and consequently the interior joints of the first storey show some slight damage. However
the overall response is consistent with what have been observed during the experimental test.

Loading Top Drift 1.2% Top Drift 1.2% Loading


Direction Direction

2.36 3.07 3.04 2.83 2.83 3.30 2.99 2.33

1.27 2.21 1.99 1.19 1.16 1.75 2.44 1.55

1.66 2.67 2.69 2.37 2.38 3.11 2.52 1.54

2.57 3.26 3.36 3.39 3.23 3.90 3.07 2.41

Figure 5.28: Damage pattern and curvature ductility at 1.2% top drift level (force controlled pushover)

- 64 -
Chapter 5 – Validation of the analytical model

Cyclic Analysis

The cyclic analysis results were compared with the hysteretic response of the test frame to
investigate the accuracy of the modeling solution in describing the non-linear cyclic behavior of an
existing frame. Furthermore a comparison between analytical and experimental results has been
made in terms of equivalent viscous damping calculated on the 1st and 3rd hysteretic cycle. Finally
the shear deformation experienced by the panel zone regions during the experimental test was
compared with the rotation of the joint springs adopted in the modeling.
As described in paragraph 5.3.4 a bilinear hysteresis rule has been adopted for the column members,
for this reason the initial stiffness obtained numerically is lower then the effective elastic stiffness of
the frame. A good general agreement in global results is shown from the figure 5.29, the
experimental response is characterized by a more marked "pinching" effect mainly due to the
splippage of longitudinal reinforcing bars of beams and columns. This effect cannot be completely
described by the Takeda hysteresis rule associated to the structural elements and the only
contribution to the "pinching" phenomenon is due to the inelastic behavior of the joint springs. A
very good agreement with the experimental results is found in terms of maximum base shear
capacity (52 kN) both in the positive and negative direction.

Top Floor Drif t (%)


-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
80

60

40
Base Shear (kN)

20

-20

-40

-60

-80
-120 -80 -40 0 40 80 120
Top Floor Displacement (mm)

Figure 5.29: Analytical-experimental comparison of the global hysteretic behavior of the frame (after Pampanin,
2002)

- 65 -
Chapter 5 – Validation of the analytical model

The values of equivalent viscous damping ξ , referred to the global Base Shear-Top Drift hysteretic
behavior, were computed for the first and third cycle at each drift level, ranging between values of ξ
= 10% and ξ = 20% (when considering cycles above the 0.6% top drift level) with a consistent
reduction (ξ values lower than 10%) when subjected to the second and third cycles at the same drift
level, due to both stiffness and strength degradation. At low level of top drift (0.2%) the numerical
model doesn't experiences any inelasticity so the correspondent equivalent viscous damping value
was taken equal to 5% to take into account the energy dissipated by the opening of cracks in the
beams and columns. At higher top drift level (0.6% and 1.2%) the value of ξ calculated from the
analysis is consistent with what obtained form the test. It is worth noting that the overestimation of
the equivalent viscous damping value corresponding to the 3rd loading cycle at 1.2% top drift is
mainly due to the incapability of accurately describe the "pinching" effect that governs the
hysteretic response of the frame.
Top Displacement (mm)
0 10 20 30 40 50 60 70 80 90
25
Equivalent Viscous Damping (%)

20
1st cycle
15

3rd cycle
10

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Top Drift (%)

Experimental
Analytical

Figure 5.30: Experimental and analytical equivalent viscous damping values (after Pampanin, 2002)

Finally the correspondence between the joint shear deformation and the spring rotation is
investigated. The exterior joint at the first and second storey are considered (figure 5.31).

- 66 -
Chapter 5 – Validation of the analytical model

LS IS IN LN

T2-S C2-S C2-N T2-N

T1-S C1-S C1-N T1-N

Figure 5.31: Test frame joint identification labels

A tendency to overestimating the inelastic deformation of the joint panel region is noticeable in
figure 5.32. It can be explained considering that the calibration of the parameters governing the
hysteretic behavior of the joint springs has been made on the basis of the global response of the
beam-column subassembly rather that on the effective shear deformation of the node region. Thus
the joint modeling approach takes into account the contribution to the rotation due to external
effects like the slippage of longitudinal steel bars in cracked beams, even if a plastic hinge is not yet
developed. Looking at the graphs it appears that the best correspondence between results has been
obtained for the south side exterior joint of the first storey (T1-S) where the model describes
satisfactorily the joint response in terms of deformation amplitude.

- 67 -
Chapter 5 – Validation of the analytical model

T1-S T1-N
0.025 0.025
Joint Shear Def ormation [rad]

0.02 0.02
0.015 0.015
0.01 0.01
0.005 0.005
0 0
-0.005 -0.005
-0.01 -0.01
-0.015 -0.015
-0.02 -0.02
-0.025 -0.025
0 1000 2000 3000 4000 5000 6000 0 1000 2000 3000 4000 5000 6000
Test Sequence Test Sequence

T2-S T2-N
0.0125 0.0125
Joint Shear Deformation [rad]

0.0075 0.0075

0.0025 0.0025

-0.0025 -0.0025

-0.0075 -0.0075

-0.0125 -0.0125
0 1000 2000 3000 4000 5000 6000 0 1000 2000 3000 4000 5000 6000
Test Sequence Test Sequence

Analytical
Experimental

Figure 5.32: Analytical and experimental joint shear deformation (after Pampanin, 2002)

- 68 -
Chapter 5 – Validation of the analytical model

5.4. REINFORCED COCRETE INFILLED FRAMES

The final stage of the model validation process was oriented to the evaluation of the accuracy of the
proposed model including equivalent-struts using the Crisafulli hysteresis rule to represent the infill
panels. The validation has been first carried out comparing the numerical results with experimental
pseudo-dynamic tests on a two dimensional one storey one bay frame performed at the Structural
Laboratory of the University of L'Aquila (Colangelo, 1999, 2003). At a second stage the
experimental results from the pseudo-dynamic tests on a four-storey, full scale, R.C. three-
dimensional building carried out at the European Laboratory for Structural Assessment (ELSA) of
the Joint Research Centre in Ispra were used as a comparative basis (Negro, 1995).

5.4.1 One storey infilled frames

The experimental investigation performed at the University of L'Aquila comprised of six infilled
one-storey 1/2 scale frames, either typical of the Italian construction practice before the introduction
of modern seismic design provisions (designed for gravity only) or designed accordingly to the
Eurocode 8 (1998). Different infill types, consisting of either vertically or horizontally hollowed
bricks, arranged in a single or double panel.
The validation of the numerical model involved the comparison with the experimental test response
of two specimens: specimen L2, designed according to the EC8 with vertically hollowed bricks
arranged in a single panel; specimen N1, designed for gravity load only following an allowable
stress approach and adopting hollow bricks arranged in a double panel.
The geometrical dimensions of the typical frame used for the tests is shown in figure 5.33. The infill
panel of the specimen L2 is composed by masonry bricks (12x25x12 cm) with vertical holes. The
N1 specimen is provided with a double panel masonry infill made of horizontal hollowed bricks
(8x25x12 cm).

- 69 -
Chapter 5 – Validation of the analytical model

a)

b)

Figure 5.33: Test frames: a) L2 specimen; b) N1 specimen (Colangelo, 2003)

Figure 5.34: Bricks adopted for the specimens L2 (left) and N1 (right) (Colangelo, 2003)

- 70 -
Chapter 5 – Validation of the analytical model

Analytical-experimental comparison

The model used in the numerical analyses follows the approach described in chapter 4. In particular
beams and columns were modeled with mono-dimensional elements with lumped inelasticity at
member ends. The cyclic behavior has been described by a modified Takeda hysteresis rule. The
joints were modeled through a rotational spring able to describe the differential rotation between
beam and column and to simulate the joint shear deformation. The joints of the L2 specimen,
designed accordingly to EC8, are represented by elastic springs. For the specimen N1, typical of
older building design, a Takeda hysteretic rule has been assigned to the rotational springs in order to
describe the post-cracking behavior of the beam-column connections. The infill panels were
modeled by equivalent diagonal struts which mechanical properties have been defined accordingly
to what described in paragraph 4.3.2. The cyclic behavior of the masonry infills has been
represented adopting the Crisafulli hysteresis rule.

- 71 -
Chapter 5 – Validation of the analytical model

Drif t [%] 0

-1

-2
0 1 2 3 4 5 6 7
Time [sec]
a)

300

200
Base Shear [kN]

100

-100

-200

-300
0 1 2 3 4 5 6 7
Time [sec]
b)

Experimental
Analytical

Figure 5.35: Analytical experimental comparison for specimen L2: a) Top drift time history; b) Base shear time
hystory (Baletta, 2002)

- 72 -
Chapter 5 – Validation of the analytical model

Drif t [%] 0

-1

-2

-3
0 1 2 3 4 5 6 7
Time [sec]
a)

200

100
Base Shear [kN]

-100

-200
0 1 2 3 4 5 6 7
Time [sec]
b)

Experimental
Analytical

Figure 5.36: Analytical experimental comparison for specimen N1: a) Top drift time history; b) Base shear time
hystory (Baletta, 2002)

The numerical – experimental results comparison show a good agreement both in terns of top drift
and base shear for the L2 specimen. In particular the correct prediction of the initial stage of the

- 73 -
Chapter 5 – Validation of the analytical model

response, governed by the masonry infill stiffness, denotes that the modeling approach allows to
adequately describe the effect of the infill panel of the overall response of the frame.
The response of the N1 specimen model is very accurate in terms of force prediction but less in
terms of displacement. It is supposed that the discrepancy in terms of top drift is mainly due to the
damage in the frame leading to slippage of longitudinal reinforcing bars in beams and columns,
typically occurring in under designed frames. The type of hysteresis rule adopted in the modeling of
the structural members is not able to represent the "pinching" behavior associated to slip
phenomena, thus underestimating the effective value of displacement experienced by the system.
More detailed information regarding the characteristics of the model and on the validation
procedure can be found in Baletta (2002).

5.4.2 Multi-storey three-dimensional infilled frames

The evaluation of the modeling approach on a more complex structure has been carried out through
a comparison with the experimental results of pseudo-dynamic tests on a full scale four storey 3D
infilled frame performed at the European Laboratory for Structural Assessment (ELSA) of the Joint
Research Centre in Ispra. Three different solution in the infills distribution along the elevation were
tested: bare frame, uniformly infilled frame and partially infilled frame (no infills at the ground
floor). The reinforced concrete frame design has been carried out in accordance with the provisions
given by the Eurocode 2 (1991) and 8 (1998) assuming a design PGA of 0.3 g and Soil Type B. The
building has a square plan, 10x10 m calculated at the columns centerline. The interstorey height is
equal to 3.5 m for the first floor and 3 m for the upper storey. The structure is composed by three
parallel frames at a distance of 5 m, each frame is made of two bays of 6 and 4 m respectively
(figure 5.37).
All the columns have a square cross section (400x400 mm) except for the central one having a
450x450 mm cross section. The beams have rectangular cross section with 450 mm depth and 300
mm width. The infills panels are constituted of vertical hollowed bricks (245x112x190 mm) with a
void ratio of 42%.
The accelerogram used for the pseudo-dynamic test was artificially generated from the EW record
of Tolmezzo (Friuli, Italy, 6/5/1976) to be compatible with the EC8 elastic design spectrum. Further
information on the specimen properties, testing procedure and experimental results can be found in
Negro (1995).

- 74 -
Chapter 5 – Validation of the analytical model

D ire c tio n o f
te s tin g

Figure 5.37: Plan and elevation view of the test frame (Negro, 1995)

Figure 5.38: Test specimen (Negro, 1995)

Analytical – experimental comparison

Regardless of the distribution of infills, the analytical-experimental comparison confirmed the


satisfactory prediction capabilities of the numerical model adopted.

- 75 -
Chapter 5 – Validation of the analytical model

3000
0.04
2000
Displacement [m]

0.02

Base Shear [kN]


1000

0 0

-0.02 -1000

-2000
-0.04
-3000
0 2 4 6 8 0 2 4 6 8
Time [sec] Time [sec]

Experimental
Analytical

Figure 5.39: Analytical-experimental comparison of the uniformly infilled frame: first floor displacement and
base shear time history (Galli, 2003)

0.15 3000

0.1 2000
Displacement [m]

Base Shear [kN]

0.05 1000

0 0

-0.05 -1000

-0.1 -2000

-0.15 -3000
0 2 4 6 8 0 2 4 6 8
Time [sec] Time [sec]

Experimental
Analytical

Figure 5.40: Analytical-experimental comparison of the partially infilled frame: first floor displacement and base
shear time history (Galli, 2003)

The figures 5.39 and 5.40 show a very good agreement between experimental and analytical results
confirming the validity of the modeling approach adopted for the representation of the infill panels.
In particular it is worth noting that the correspondence between the results is adequate at any stage
of the analysis denoting a realistic evaluation of the elastic properties of the equivalent diagonal
struts as well as of the masonry cyclic behavior. More detailed information about the analytical –
experimental comparisons can be found in Galli (2003).

- 76 -
Chapter 6 – Numerical investigation on 2D frame systems

Equation Section (Next)

6. NUMERICAL INVESTIGATION ON 2D FRAME SYSTEMS

The process of calibration and validation of the numerical model based on a concentrated plasticity
approach leaded to the development of a simple and, at the same time, reliable analytical tool. This
modeling approach is a valuable support for extensive numerical investigation studies that are one
of the scopes of the present work. In particular, the main object of the study is the analysis of the
two and three dimensional seismic response of existing frame buildings designed for gravity load
only, as commonly done in the Italian engineering practice of the past. Another interesting topic to
be investigated through dynamic time history analyses is the effect of the presence of masonry
infills in the frame; different solution in the distribution of infill panels are considered: bare frame,
uniformly infilled frames with strong and weak panels, to represent external walls and internal
partitions respectively, and partially infilled frames (without infills at the ground floor).
In this chapter the attention is focused of the analysis of 2D frames, first analyzing the 2/3 scaled
three storey frame tested at the Department of Structural Mechanics of the University of Pavia and
described in the previous chapter. At a second stage the seismic response of a case-study 6 storey
three-bays frame system has been analyzed first performing some pushover analyses and then
through a series of non-linear dynamic time history analyses.

- 77 -
Chapter 6 – Numerical investigation on 2D frame systems

6.1. THREE STOREY FRAME

The numerical investigation carried out in the following paragraphs is focused on the evaluation of
the seismic response of a 2/3 scaled three-storey three-bays reinforced concrete frame tested at the
University of Pavia. The test frame has been described in paragraph 5.3. The model of the bare
frame has been then modified by the introduction of equivalent diagonal struts to simulate the
presence of masonry infills accordingly to two widely adopted structural schemes: a) uniform
distribution of infills; b) partial distribution of infills (no infills at the ground level).

6.1.1 Description of the infilled model

Figure 6.1 shows the distributions of masonry infills adopted for the numerical investigations. The
two solutions are considered as the most representative for a wide range of existing buildings built
between 1950 and 1970 in the Mediterranean area.

a) b)

Figure 6.1: Three storey frame infills distribution: a) uniformly: b) partially

In both cases the frames have been studied considering masonry infills composed by a single panel
and a double panel of bricks. The model provided with weak masonry infills (single panel) can be
considered representative of internal partition walls used to separate different rooms inside the
building. The strong infills (double panel), instead, have been chosen to model the external walls
which are often built using two parallel masonry panels with an air inter-space in between, with
insulating purpose.
The properties of the equivalent diagonal struts used in the model were defined to be representative
of a masonry type similar to the one used for the pseudo-dynamic tests performed at the Structural
Laboratory of the University of L'Aquila (Colangelo, 1999). The wallets tested were composed of
- 78 -
Chapter 6 – Numerical investigation on 2D frame systems

horizontal hollowed bricks (246x118x79 mm), with a void ratio equal to 53.8%, a compressive
strength equal to 16.36 MPa in the direction parallel to the holes and 2.19 MPa in the perpendicular
direction. The wall specimens were tested under compression load parallel and perpendicular to the
holes and under diagonal compression. The mean values of the mechanical quantities that
characterize the masonry are reported in table 6.1 together with the correspondent coefficient of
variation (c.o.v.).

Quantity Mean c.o.v.


fwh [MPa] 3.84 24.9%
fwv [MPa] 2.7 3.7%
Ewh [MPa] 2586 66.8%
Ewv [MPa] 1195 44%
G [MPa] 1389 6.8%
fws [MPa] 0.57 3.1%
fwu [MPa] 0.3 -
ν 0.2 -
Table 6.1: Masonry mechanical properties: mean value and c.o.v

with:
fwh horizontal compressive strength (parallel to the holes)
fwv vertical compressive strength (perpendicular to the holes)
Ewh horizontal elastic modulus of masonry
Ewv vertical elastic modulus of masonry
G shear modulus of masonry obtained from diagonal compression test
fws shear strength of masonry obtained from diagonal compression test
fwu sliding resistance of mortar
ν Poisson coefficient
The values of fws and ν has been taken as representative of an average masonry since any indication
was available in the experimental documentation.
In worth noting how the values of the mechanical properties of masonry are highly affected by
dispersion respect to the mean values, thus indicating how difficult is to model adequately the
masonry infills characteristics and their effect on the global response of the structure in which they
are inserted.
The stiffness and strength of the equivalent diagonal struts have been evaluated accordingly to what
described in paragraph 4.3.2. In particular the width of the strut cross section has been evaluated
bw K1
with the formula: = + K 2 (Decanini et al., 1993) where the constants K1 and K2 can be read
dw λh

- 79 -
Chapter 6 – Numerical investigation on 2D frame systems

in table 4.1. The elastic modulus of the masonry in the direction of the strut, inclined at an angle α
respect to the horizontal, has been evaluated with the equation 4.7.
The strength assigned to the strut member is the values correspondent to the most probable failure
mechanism among for possible mechanisms taken into account: compression at centre of panel,
compression of corners, sliding shear and diagonal tension.

6.1.2 Pushover analyses of infilled frames

A series of pushover analyses has been performed on the uniformly and partially infilled frame,
considering both the single and the double panel solution. A triangular distribution of forces has
been applied to the structure similarly to what have been done for the bare frame. In particular the
ratio between the storey forces is shown in table 5.12.

Top Displacement [mm]


0 30 60 90
250

Uniformly Infilled (2 panel)


200
Uniformly Infilled (1 panel)
Base Shear [kN]

150

100

50

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.2: Base shear-top drift curve of the uniformly infilled frame

Comparing the monotonic response of the uniformly infilled frames to the bare frame response the
contribution of infill panels in terms of strength and stiffness appear evident. As expected the single
panel infilled frame (dashed line curve) shows a lower value of strength and after the masonry
infills have cracked the response of the two frames converge to the same value of force,
corresponding to the response curve of the bare frame.

- 80 -
Chapter 6 – Numerical investigation on 2D frame systems

Top Drift [%]


0 0.5 1 1.5

3 6

Height [m]
2 4

Storey
1 2

0 0
0 20 40 60 80
Top Displacement [mm]

Double Panel Infill


Single Panel Infill

Figure 6.3: Displacement profile of the uniformly infilled frame

The displacement profile of the uniformly infilled frames analyzed clearly shows the development
of a soft storey mechanism located at the first level. The infill panels start to damage at
displacement value correspondent to a top drift level of 0.5%. The inelastic deformation request at
the first storey rapidly increases with loading and mainly all the deformation of the structure is
concentrated at this level. The response, in terms of displacement, of the frames infilled with single
and double masonry panels appears very similar and characterized by a storey mechanism at the
lower level.
The pushover analysis performed on the partially infilled frame shows a monotonic behavior of the
building that is very similar to the response of the bare frame (figure 6.4). It is worth noting that,
similarly to what happens to the bare frame, the response of the partially infilled frame is dominated
by the inelastic deformation of base columns where the absence of infill panels leads to a drastic
reduction of stiffness compared with the upper storey. The response of the structure is not
influenced by the type of masonry infill used since the infills suffer very little damage and the soft
storey mechanism develops at the first storey where no infill is present. The displacement profile
plot (figure 6.5) clearly shows the type of mechanism governing the global response of the partially
infilled frame, regardless of the type of masonry panels present.

- 81 -
Chapter 6 – Numerical investigation on 2D frame systems

Top Displacement [mm]


0 30 60 90
80

60
Base Shear [kN]

40

Panrtially Infilled (2 panel)


20
Partially Infilled (1 panel)

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.4: Base shear-top drift curve of the partially infilled frame

Top Drift [%]


0 0.5 1 1.5 2

3 6
Height [m]

2 4
Storey

1 2

0 0
0 40 80 120
Top Displacement [mm]

Double Panel Infill


Single Panel Infill

Figure 6.5: Displacement profile of the partially infilled frame

- 82 -
Chapter 6 – Numerical investigation on 2D frame systems

Top Displacement [mm]


0 30 60 90
200

Bare Frame
160 Uniformly Infilled Frame
Partial Infilled Frame
Base Shear [kN]

120

80

40

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.6: Comparison between the 3 storey frame pushover curves

Figure 6.6 shows a comparison between the pushover curves of the three typologies of frame
analyzed. It is worth noting how the presence of masonry infills increases the strength and stiffness
of the structure. The bare frame curve shows a higher initial stiffness because the columns were
modeled with a Fukada tri-linear hysteresis rule, taking into account the uncracked response. For
this reason the first stiffness of the bare frame is similar to the initial stiffness of the partial infilled
frame. The uniform infilled frame, instead, show a response characterized by higher strength and
stiffness up to the failure point of the masonry panels and the consequent development of the storey
mechanism. In the post elastic range the curve rapidly decreases eventually falling on the response
curve of the partial infilled frame

- 83 -
Chapter 6 – Numerical investigation on 2D frame systems

6.1.3 Records used for the time history analyses

The time history analyses have been performed with the dynamic analysis program Ruaumoko 2D
(Carr, 2004) using ten recorded accelerograms chosen between a more complete set of Californian
input ground motions used in other studies on seismic assessment of frame systems (Pampanin,
2003). The characteristics of the accelerograms as taken from the original set are shown in table 6.2.

EQ n° Event Year Magnitude PGA [g] Duration [sec]


EQ1 Cape Mendocino 1992 7.1 0.44 44
EQ2 Landers 1992 7.3 0.42 50
EQ3 Landers 1992 7.3 0.33 44
EQ4 Loma Prieta 1989 6.9 0.36 40
EQ5 Northridge 1994 6.7 0.37 30
EQ6 Northridge 1994 6.7 0.43 25
EQ7 Northridge 1994 6.7 0.44 40
EQ8 Northridge 1994 6.7 0.35 30
EQ9 Superstition Hills 1987 6.7 0.49 40
EQ10 Superstition Hills 1987 6.7 0.41 22
Table 6.2: Record used for the time history analyses

The records have been scaled to 75% of the original PGA value so that the average spectrum was
compatible with the EC8 elastic spectrum at 5% damping, considering PGA = 0.3g and soil type B.
1
Mean elastic spectrum
EC8 elastic spectrum
0.8

0.6
PGA [g]

0.4

0.2

0
0 1 2 3 4 5
Period [sec]

Figure 6.7: Comparison between EC8 elastic spectrum (PGA = 0.3g; Soil type B) and mean response spectrum of
ten records

- 84 -
Chapter 6 – Numerical investigation on 2D frame systems

For the analysis of the 2/3 scaled reinforced concrete frame the records were scaled so that the
duration and the amplitude of the accelerograms were consistent with the other physical quantities
of the scaled specimen. In particular the original acceleration has been divided by the scale factor β
= 2/3, thus obtaining an higher value of the acceleration to be used for the analysis of the scaled
model. On the other hand the duration of the records is reduced multiplying the real time length by
the scale factor. Table 6.3 reports the coefficients to be used for the main physical quantities to
obtain the correspondent value when a model, scaled by a factor β, is analyzed. The following
relationships are obtained with the hypothesis of keeping constant the material density and the
stress.

Quantity Prototype Model


Length l βl
Mass M β3M
Time t βt
Stress s s
Velocity v v
Acceleration a β-1a
Force F β2F
Damping c β2c
Table 6.3: Scale factors for the principal physical quantities

6.1.4 Time history results

In the following pages the average results obtained from the analyses using 10 input records are
reported, together with some more detailed results for three particular accelerograms: Loma Prieta
(EQ4); Northridge (EQ8); Superstition Hills (EQ10).

Bare frame

The results of the dynamic time history analyses performed of the bare frame model are shown in
figure 6.8 and reported in table 6.4 as average of the ten ground motions. The plots show the
maximum displacement and maximum drift profile recorded during the analysis to have an
indication of the maximum deformation experienced by the frame. Residual values of displacement
and drift are also reported as a good indicator of the damage suffered by the structure.

- 85 -
Chapter 6 – Numerical investigation on 2D frame systems

Maximum displacement [m] Drift [%]


storey negative positive residual maximum residual
1 -0.055 (0.028) 0.046 (0.020) -0.0042 3.20 (1.33) 0.44
2 -0.067 (0.028) 0.059 (0.020) -0.0037 0.79 (0.08) 0.05
3 -0.071 (0.028) 0.063 (0.020) -0.0036 0.38 (0.07) 0.01
Table 6.4: Displacement and interstorey drift: mean values and standard deviation of ten analyses on the bare
frame

3 3 6

Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

3 3 6

Height (m)
Storey

2 4
Storey

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2

Residual Interstorey Drift (%) Residual Displacement (m)

Figure 6.8: Time-history analysis on the 3-storey bare frame: interstorey drift and displacement profile
(average of maximum and residual)

Looking at the plot of the average maximum drift profile it appear evident a high request of inelastic
deformation at the first storey where a soft storey mechanism is likely to develop. The plastic
deformation and consequent probable damage suffered by the elements of the first storey is also
evidenced by an average residual drift value of 0.44%.
In the following pages three records (Loma Prieta (EQ4); Northridge (EQ8); Superstition Hills
(EQ10)) are analyzed in more details. In particular the level reached by the joint deformation is
investigated together with the contribution of the different elements to the maximum drift. Figures
6.9 to 6.11 show the global results for the three storey frame.

- 86 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


Top Floor Displacement [m] 1.5
0.08
1

Top Drift [%]


0.04
0.5
0 0
-0.5
-0.04
-1
-0.08
-1.5

80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25 30 35 40
Time [sec]

3 3 6

Height (m)
Storey

2 4
Storey

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drift [%] Envelope Max Floor Displacements [m]

Figure 6.9: Bare frame: time history analysis results (Loma Prieta, EQ4)

- 87 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
Top Floor Displacement [m]
0.12 2
0.08

Top Drif t [%]


1
0.04
0 0
-0.04
-1
-0.08
-0.12 -2

80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25 30
Time [sec]

3 3 6

Height (m)
Storey

2 4
Storey

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drift [%] Envelope Max Floor Displacements [m]
Figure 6.10: Bare frame: time history analysis results (Northridge, EQ8)

- 88 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills(EQ10)
0.08
Top Floor Displacement [m]
1
0.04

Top Drift [%]


0.5

0 0

-0.5
-0.04
-1
-0.08

80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25
Time [sec]

3 3 6

Height (m)
Storey

2 4
Storey

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drift [%] Envelope Max Floor Displacements [m]

Figure 6.11: Bare frame: time history analysis results (Superstition Hills, EQ10)

- 89 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


3 3 3

2 2 2
Storey N°

Storey N°

Storey N°
1 1 1

0 0 0
0 0.004 0.008 0 0.004 0.008 0 0.004 0.008
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]

Column 1
Column 2
Column 3
Column 4

Figure 6.12: Bare frame joint rotations

Figure 6.12 shows the maximum joint rotation recorded during the dynamic analyses with the three
records under consideration. All the plots indicate a higher level of deformation for the exterior
joints (column 1 and 4), this result confirms what has been obtained from the experimental tests on
the beam-column subassemblies and on the three storey frame tested at the University of Pavia.
This behavior has been also predicted by the pushover analyses performed on the three storey R.C,
frame model (par. 5.3). The maximum and residual displacement profiles for the EQ4 and EQ8
records show a marked difference between positive and negative direction and this is confirmed by
the joint rotations which are higher on one side of the frame respect to the other, indicating a more
severe level of damage reached by one particular node.
Figures 6.13 to 6.15 report the contribution of each single element (beam, column and joint) of the
beam column connection to the maximum interstorey drift. These plots show again that the higher
contribution to the drift comes from the exterior joints located at the first storey level.
However it is worth noting that the joint rotation never exceeds the value of 0.01 rad, that can be
assumed as a limit state for critical damage (Pampanin, 2002).

- 90 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Loma Prieta (EQ4)
Column 3
Column 4

3 3

Beam
Beam
Storey N°

Storey N°
Column
2 Column 2

Joint
Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Joint

Beam
Beam
Storey N°

Storey N°

2 Column 2 Column

Joint

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.13: Bare frame: elements contribution to the maximum interstorey drift (EQ4)

- 91 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Northridge (EQ8)
Column 3
Column 4

3 3

Column
Storey N°

Storey N°
Beam

Beam
2 2

Column
Joint
Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Joint
Beam
Storey N°

Storey N°

Beam
2 2

Column Column
Joint

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.14: Bare frame: elements contribution to the maximum interstorey drift (EQ8)

- 92 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Superstition Hills (EQ10) Column 2
Column 3
Column 4

3 3

Beam

Beam
Storey N°

Storey N°
2 2
Column Column

Joint
Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 Joint 3

Beam
Beam
Storey N°

Storey N°

2 2

Column Column

Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.15: Bare frame: elements contribution to the maximum interstorey drift (EQ10)

- 93 -
Chapter 6 – Numerical investigation on 2D frame systems

Partially infilled frame

Table 6.5 reports the average of the maximum displacement recorded during the time history
analyses using ten accelerograms; the coefficient of variation, in brackets, give an indication of the
dispersion of the results depending on the type of record used as input. Table 6.6 shows the
maximum interstorey drift, and the corresponding standard deviation, reached by the structure
during the time histories. Looking at the interstorey drift values it is immediate to notice that the
inelastic deformation is nearly completely concentrated at the first storey (3.38% at the first level
versus 0.03% and 0.01% at the upper levels for the double panel infilled frame). As expected the
behavior of the partially infilled frame is governed by a soft storey mechanism located at the ground
floor, where the absence of infill panels causes a big discontinuity in terms of strength and stiffness.
The residual drift values confirm the activation of the aforementioned mechanism and indicate the
concentration of damage in the structural elements located at the first storey.

Maximum displacement [m]


Double panel infill Single panel infill
storey Negative positive residual negative positive residual
1 -0.055 (0.021) 0.046 (0.025) -0.0042 -0.047 (0.016) 0.050 (0.020) -0.0004
2 -0.067 (0.021) 0.059 (0.025) -0.0037 -0.048 (0.016) 0.051 (0.020) -0.0004
3 -0.071 (0.021) 0.063 (0.025) -0.0036 -0.049 (0.016) 0.051 (0.020) -0.0004
Table 6.5: Displacement: mean values and standard deviation of ten analyses on the partially infilled frame

Drift [%]
Double panel infill Single panel infill
storey maximum residual maximum residual
1 3.38 (1.17) 0.430 3.00 (0.79) 0.357
2 0.03 (0.00) 0.001 0.05 (0.00) 0.003
3 0.01 (0.00) 0.000 0.02 (0.00) 0.000
Table 6.6: Interstorey drift: mean values and standard deviation of ten analyses on the partially infilled frame

The graphical representation of the values shown in the previous tables is presented in figure 6.16.
Again it appear evident the storey mechanism activated at the first level where a remarkable request
of plastic deformation is located, leading to a high level of damage mainly in columns and joints.
Moreover it is worth noting how the response of the structure is not to much influenced by the type
of infills adopted (i.e. double or single panel). The absence of infills at the ground level, in fact,
causes a big stiffness discontinuity, that governs the structural behavior, regardless of level of
strength and rigidity of the masonry panels present at the upper levels.

- 94 -
Chapter 6 – Numerical investigation on 2D frame systems

3 3 6

Height [m]
Storey

Storey
2 2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

3 3 6

Height (m)
Storey

2
Storey
2 4

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Residual Interstorey Drift (%) Residual Displacement (m)

Double panel
Single panel

Figure 6.16: Time-history analysis on the 3-storey partially infilled frame: interstorey drift and displacement
profile (average of maximum and residual)

In the following pages the results of the analyses performed using as input Loma Prieta (EQ4),
Northridge (EQ8) and Superstition Hills (EQ10) records are overviewed focusing the attention on
the behavior of joints, looking at the maximum rotations experienced by the single spring, and on
the contribution of the different elements to the global interstorey drift.
In figures 6.17 to 6.19 displacement and drift profiles (maximum and residual) are reported together
with the top floor displacement and base shear time history plots. In the graphs the solid line
represents the response of the frame provided with strong infills (double panel) and the dashed line
corresponds the frame infilled with weak panels (single panel infills).
The limited values of residual displacements obtained can be explained considering that the
analyses performed do not take into account strength degradation and P-∆ effects

- 95 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


1
Top Floor Displacement [m]

0.04
0.5

Top Drift [%]


0 0

-0.5
-0.04

-1
80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25 30 35
Time [sec]

3 3 6

Height (m)
Storey

2 4
Storey

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05

Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey

2 4
Storey

1 1 2

0 0 0
0 2 4 6 -0.1 0 0.1
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.17: Partially infilled frame: time history analysis results (Loma prieta, EQ4)

- 96 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
0.06 1
Top Floor Displacement [m]

0.04
0.5

Top Drift [%]


0.02

0 0

-0.02
-0.5
-0.04

-0.06 -1

80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25 30
Time [sec]

3 3 6

Height (m)
Storey

2
Storey

2 4

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey

2 4
Storey

1 1 2

0 0 0
0 2 4 6 -0.1 0 0.1
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.18: Partially infilled frame: time history analysis results (Northridge, EQ8)

- 97 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills (EQ10)


0.06 1
Top Floor Displacement [m]

0.03 0.5

Top Drif t [%]


0 0

-0.03 -0.5

-0.06 -1

80
60
40
Base Shear [kN]

20
0
-20
-40
-60
-80
0 5 10 15 20 25
Time [sec]

3 3 6

Height (m)
Storey

2
Storey

2 4

1 1 2

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey

2 4
Storey

1 1 2

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.19: Partially infilled frame: time history analysis results (Superstition Hills, EQ10)

- 98 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


3 3 3

2 2 2

Storey N°
Storey N°
Storey N°

1 1 1

0 0 0
0 8e-005 0.00016 0 0.0004 0.0008 0.0012 0 8e-005 0.00016
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]
a) b) c)
Column 1
Column 2
Column 3
Column 4

Figure 6.20: Partially infilled frame joint rotations

As expected the higher level of joint deformation is experienced by the joints of the first storey. It
appears that there is no big difference in the value of rotation between interior and exterior joints
except for the case of the double panel infilled frame (solid line) analyzed using Northridge record
as input, where the exterior joints suffered a maximum deformation much higher than the other
nodes. It is worth noting however that generally the maximum values of rotation experienced by the
joints of the partially infilled frame are markedly lower that those of the bare frame (see figure
6.12). This can be due to the presence of the equivalent diagonal strut, representing the infill panel
and converging into the node, that exerts a stiffening effect on the beam-column subassembly so
that the deformation demand of the region is concentrated into the plastic regions lower columns.
The drift contribution plots (figure 6.21 to 6.23) confirm the predominant role of the column
members to the maximum storey deformation, especially at the first storey level where the higher
level of drift is concentrated. The following graphs are referred to the case of double panel infilled
frame that is believed to be more interesting since, respect to the single panel solution, is the
structural typology that suffered a higher level of inelastic deformation.

- 99 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Loma Prieta (EQ4)
Column 3
Column 4

3 3
Joint Joint

Beam

Beam
Storey N°

Storey N°
Column
2 Column 2

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 Beam 3
Joint
Storey N°

Storey N°

Joint

Beam
2 Column 2 Column

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.21: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ4)

- 100 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Northridge (EQ8) Column 2
Column 3
Column 4

3 3

Column

Beam
Storey N°

Storey N°

Beam
2 2 Joint

Column

Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Joint
Beam
Storey N°

Storey N°

Beam
2 2

Column Column
Joint

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]
Figure 6.22: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ8)

- 101 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Superstition Hills (EQ10) Column 3
Column 4

3 3

Beam
Storey N°

Storey N°
2 Beam 2
Column Column

Joint
Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Beam
Storey N°

Storey N°

Beam
2 2

Column Joint Column

Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]
Figure 6.23: Partially infilled frame: elements contribution to the maximum interstorey drift (EQ10)

- 102 -
Chapter 6 – Numerical investigation on 2D frame systems

Uniformly infilled frame

In the following tables the results in terms of displacement and interstorey drift of the analyses
performed of the uniformly infilled 3 storey frame are reported. The values shown are the average
of ten analyses and the values in bracket are the coefficient of variation of the correspondent
quantity.
The displacement and drift profiles of the frame infilled with strong panels (double layer of bricks)
indicate a low level of deformation (maximum interstorey drift of 0.3%) and negligible value of
residual deformation thus indicating that nearly any damage has been suffered by the structure. It is
reasonable to assume that due to the limited height and mass of the frame the stiffening effect of the
masonry panels is enough to prevent the failure of the infills and the following damage of the
structural elements. The frame provided with weak infills (single layer of bricks), instead, reached
approximately twice the value of maximum storey displacement and drift. In particular the higher
value of deformation is recorded at the first storey, where, after the failure of the masonry infills, a
soft storey mechanism is likely to develop as a consequence of a drastic and sudden decrease in
lateral stiffness.

Displacement [m]
Double panel infill Single panel infill
storey Negative positive residual negative Positive residual
1 -0.005 (0.002) 0.006 (0.001) -0.0002 -0.011 (0.004) 0.012 (0.003) -0.0001
2 -0.007 (0.002) 0.008 (0.002) -0.0001 -0.014 (0.005) 0.016 (0.003) 0.0000
3 -0.009 (0.002) 0.009 (0.002) -0.0001 -0.016 (0.005) 0.017 (0.003) 0.0001
Table 6.7: Displacement: mean values and standard deviation of ten analyses on the uniformly infilled frame

Drift [%]
Double panel infill Single panel infill
storey maximum residual Maximum residual
1 0.29 (0.07) 0.017 0.59 (0.19) 0.047
2 0.14 (0.03) 0.005 0.20 (0.05) 0.007
3 0.06 (0.02) 0.001 0.08 (0.03) 0.003
Table 6.8: Interstorey drift: mean values and standard deviation of ten analyses on the uniformly infilled frame

- 103 -
Chapter 6 – Numerical investigation on 2D frame systems

3 3 6

Height [m]
Storey

Storey
2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.03 -0.015 0 0.015 0.03
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

3 3 6

Height (m)
Storey

2
Storey
2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.03 -0.015 0 0.015 0.03
Residual Interstorey Drift (%) Residual Displacement (m)

Double panel
Single panel

Figure 6.24: Time-history analysis on the 3-storey uniformly infilled frame: interstorey drift and displacement
profile (average of maximum and residual)

Figures 6.25 to 6.27 show the top storey displacement and base shear time history for three input
records: Loma Prieta (EQ4), Northridge (EQ8) and Superstition Hills (EQ10). The plots of
maximum and residual displacement and interstorey drift versus the building height are also
presented. In the following pages the maximum joint rotations and the contribution of different
elements to the interstorey drift will be discussed.
In the following plots the dashed line is representative of the single panel infilled frame and the
solid line represents the response of the frame infilled with strong panels.

- 104 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


0.02
Top Floor Displacement [m] 0.3

0.01 0.15

Top Drift [%]


0 0

-0.01 -0.15

-0.3
-0.02

300

200
Base Shear [kN]

100

-100

-200

-300
0 5 10 15 20 25 30 35
Time [sec]

3 3 6

Height (m)
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.25: Uniformly infilled frame: time history analysis results (Loma Prieta, EQ4)

- 105 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
0.02
Top Floor Displacement [m] 0.3

0.01 0.15

Top Drift [%]


0 0

-0.01 -0.15

-0.3
-0.02

200
Base Shear [kN]

100

-100

-200

0 5 10 15 20 25 30
Time [sec]

3 3 6

Height (m)
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.26: Uniformly infilled frame: time history analysis results (Northridge, EQ8)

- 106 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills (EQ10)


0.03 0.5
Top Floor Displacement [m]
0.02
0.25

Top Drift [%]


0.01

0 0

-0.01
-0.25
-0.02

-0.03 -0.5

200
Base Shear [kN]

100

-100

-200

0 5 10 15 20 25
Time [sec]

3 3 6

Height (m)
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

3 3 6
Height [m]
Storey
Storey

2 2 4

1 1 2

0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.27: Uniformly infilled frame: time history analysis results (Superstition Hills, EQ10)

- 107 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


3 3 3

2 2 2
Storey N°

Storey N°
Storey N°
1 1 1

0 0 0
0 8e-005 0.00016 0 0.0004 0.0008 0.0012 0 0.0004 0.0008
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]
a) b) c)

Column 1
Column 2
Column 3
Column 4

Figure 6.28: Uniformly infilled frame joint rotations

As noticed in the cases of bare and partially infilled frame, the joints that suffered the higher values
of rotation are the exterior nodes of the first storey. Looking at the graphs in figure 6.28 it is evident
that the joints of the frame with strong infills experience a low level of shear deformation as a
consequence of the overall limited value of deformation reached by the structure. On the other hand
it is evident how the failure of the weak infills (dashed lines) lead to a more severe request of
inelastic deformation request on the beam-column subassemblies. In some cases (see fig 6.28 b) the
joint rotations exceed the value of 0.001 rad that can be assumed as a limit state of critical damage
for the beam column connection. In this case the extensive damage of the node region could
compromise the bearing capacity of the frame or at least require big efforts in repairing the
structure.
The following plots show the important contribution of the columns deformation to the maximum
interstorey drift. Moreover they confirm the greater contribute of the exterior joints to the frame
deformation, as already noticed in the analyses on the bare and infilled frames.

- 108 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Loma Prieta (EQ4) Column 3
Column 4

3 3
Joint Joint

Beam
Storey N°

Storey N°
Beam

Column
2 Column 2

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Joint Joint
Storey N°

Storey N°

Beam
Beam
2 Column 2 Column

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.29: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ4)

- 109 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Northridge (EQ8) Column 3
Column 4

3 3

Column Beam
Storey N°

Storey N°

Beam
2 2 Joint

Column

Joint
1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 3
Joint
Beam
Storey N°

Storey N°

Beam
2 2

Column Column
Joint

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.30: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ8)

- 110 -
Chapter 6 – Numerical investigation on 2D frame systems

Column 1
Column 2
Superstition Hills (EQ10) Column 3
Column 4

3 3 Joint

Beam
Storey N°

Storey N°
Beam

2 2
Column Column

Joint

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

3 Joint 3

Joint
Storey N°

Storey N°
Beam

Beam
2 2

Column Column

1 1
0 20 40 60 80 100 0 20 40 60 80 100
Drift contribution [%] Drift contribution [%]

Figure 6.31: Uniformly infilled frame: elements contribution to the maximum interstorey drift (EQ10)

- 111 -
Chapter 6 – Numerical investigation on 2D frame systems

6.2. SIX STOREY FRAME

After the investigation on the seismic response of the three storey frame has been carried out, the
modeling procedure described in the previous chapters has been applied to a case-study six-storey
three-bay frame system assumed as typical of the buildings designed in Italy between the 1950 and
1970. The structure as been studied in three different configurations: bare frame, uniformly infilled
frame (uniform distribution of infills along the elevation) and non-uniformly infilled frame ( no
infills at the first storey). Either single or double panel arrangements for the infills were considered.
In the following paragraphs the case-study frame is described and the results from non-linear
pushover and time history analyses are presented.

C1 C1 C1 C1
3m

C1 C1 C1 C1
3m

C1 C1 C1 C1
3m
17,75 m

3m

C1 C2 C2 C1
3m

C1 C2 C2 C1
2,75 m

C2 C3 C3 C2

4,5 m 2m 4,5 m

11 m

Figure 6.32: Geometric dimensions of the 6-storey frame

- 112 -
Chapter 6 – Numerical investigation on 2D frame systems

6.2.1 Description of the frame

The structure is considered as part of a frame system building formed by a series of parallel frames
at a distance of 4.5 m between centerlines of columns. The frame is six storey high with a
interstorey height of 3 m with the exception of the first storey (2.75 m), and is constituted by three
bays. The middle span is 2 m long while the exterior bays are 4.5 m long.
The design of the frame as been carried out assuming the materials usually adopted in the 50's:
• plain round bars with characteristic yielding strength fy = 3800 Kg/cm2 and allowable
stress σs = 1600 Kg/cm2
• concrete with characteristic cubic compressive strength Rck = 200 Kg/cm2 .

The value of allowable stress for concrete was evaluated with the following relationship:
Rck − 150
σ c = 60 + = 72.5 Kg/cm 2
4

The structural elements has been designed following the allowable stress approach, accordingly to
the Italian Code Provisions and design hand-books available at the time (1950-70). The beams were
designed considering a continuous beam scheme on four supports and the accidental design load
was taken as 600 Kg/m2 on the floor slabs and 500 Kg/m2 on the roof.
The geometrical dimensions of beams were takes as 300 mm (width) and 500 mm (depth) and the
required amount of reinforcement steel has been evaluated with the following equation:
M
As = (6.1)
0.9dσ s

where M is the design moment, d is the effective depth of the section and σs is the steel allowable
stress.
The column dimensions were obtained considering the element simply subjected to the design axial
compression force and using the following relation:
N
Ac = (6.2)
σc
where σc is the allowable compressive strength of concrete.
As a common design practice between 50's and 70's the amount of longitudinal reinforcing steel has
been evaluated imposing a reinforcing ratio ρ = 1%.

- 113 -
Chapter 6 – Numerical investigation on 2D frame systems

A B A A B A A B A

A B A A B A A B A

4,5 m 2m 4,5 m
11 m

Φ 6φ6/10cm
staffe / 10cm Φ 6φ6/20cm
staffe / 10cm Φ 6 /φ6/10cm
staffe 10cm Φ 6φ6/20cm
staffe / 10cm Φ 6φ6/10cm
staffe / 10cm Φ 6 φ6/20cm
staffe / 10cm Φ 6 /φ6/10cm
staffe 10cm

2 φ16

2 φ16

2 φ16

Section
SezioneA-A
A-A Section
SezioneB-B
B-B
4 φ16 2 φ16

20 20

2 φ16 4 φ16
500

500

300 300

Figure 6.33: Beam sections: geometric and mechanical characteristics

PilastroC1
Section C1 PilastroC2
Section C2 Pilastro C3
Section C3

20
4 φ16 20
4 φ16
20
250

4 φ18
300

350

250 300
350

Figure 6.34: Column sections of the 6-storey frame

Longitudinal reinforcement Stirrups diameter and Cover


[mm] spacing [mm] [mm]
C1 4Φ16 Φ6 / 100 20
C2 4Φ16 Φ6 / 150 20
C3 4Φ18 Φ6 / 150 20

Table 6.9: Column reinforcement summary table

- 114 -
Chapter 6 – Numerical investigation on 2D frame systems

6.2.2 Frame modeling

The modeling of beams and columns has been made through members with lumped plasticity at the
ends. In this way the cyclic behavior of elements is governed by rotational springs, located in the
sections were a plastic hinge is more likely to occur. To describe the inelastic behavior of beam and
column members a modified Takeda (Otani, 1974) hysteresis rule has been adopted. The parameters
which define the loop (α and β) have been chosen in order to minimize the energy dissipation
capacity because the structures designed for gravity loads only are usually characterized by a low
dissipative behavior. In particular the following values has been used:
• a = 0.5 to define the unloading stiffness

• b = 0.0 to define the reloading point on the monotonic curve

For beams, since the hysteretic rule adopted doesn't allow to define different values for the positive
and negative stiffness, an average value for the cracked stiffness in both direction has been
assumed.
The beam-column joints has been modeled, as described in paragraph 4.2.2 and chapter 5, with a
couple of axial-rotational spring to which a M-N interaction relationship has been associated. This
type of modeling allows to describe the relative rotation between beams and columns converging
into the node and the post-cracking shear deformation of the panel region. Furthermore the presence
of a moment-axial load interaction diagram associated to the spring elements allows to take into
account the affect of the variation of axial load, typical of structures subjected to seismic loading,
on the shear joint capacity. The joint elements have been provided with the hysteretic loop
described in paragraph 5.2.3 and capable to represent the "pinching" effect typical of the behavior
of beam-column joints in existing buildings and mainly due to the slippage of longitudinal
reinforcing bars and the shear strength degradation, result of the joint cracking. The parameters used
to define the hysteretic cycle has been deduced to the validation process on beam-column
subassemblies described in paragraph 5.2.3.
The frame has been analyzed with different distribution of infill panels (uniformly infilled and
partially infilled) and with different typology of infills (strong panels with double layer of bricks
and weak panels with a single layer of bricks). The weak infill panels are considered representative
of interior frames were the walls are used as partitions between rooms. Strong infills instead can
represent exterior frame of the building were the panels are used as perimeter walls of the structure.
Equivalent diagonal struts have been used to model the infill panels and the properties of the struts

- 115 -
Chapter 6 – Numerical investigation on 2D frame systems

have been calculated on the basis of the mechanical properties of wallets tested at the University of
L'Aquila (Colangelo, 1999) (see table 6.1).

6.2.3 Pushover analysis

A series of pushover analyses has been performed on the six storey frames, considering all the
different infills distribution options (bare, partially infilled and uniformly infilled frames). A
triangular force distribution has been assumed consistently with what have been done for the 3
storey frame system. The applied lateral forces has been calculated with the following equation:
mi ⋅ hi
Fi = ⋅ Fbase (6.3)
∑ r r
r
m ⋅ h

with:
Fi is the force at the ith storey
mi is the mass at the ith storey
hi is the height of the ith storey
Fbase is the total base shear

Bare frame
Top Displacement [m]
0 0.1 0.2

300
Base Shear [kN]

200

100

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.35: Pushover analysis on the 6-storey bare frame: Base shear - Top drift curve

- 116 -
Chapter 6 – Numerical investigation on 2D frame systems

Top Drift [%]


0 0.5 1 1.5

6 6
16 16

Height [m]
12

Height [m]
4 12 4

Storey
Storey

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 2 4 6
Top Displacement [m] Interstorey Drift [%]

Figure 6.36: Pushover analysis on the 6-storey bare frame: Displacements and Interstorey drift profiles
corresponding to top drift value of 0.5% and 1%.

The pushover analysis performed on the bare frame system shows an evident soft storey mechanism
developing between the 3re and 4th storey. This behavior could be expected considering that at this
particular level there is a sensible reduction in the columns cross sections thus leading to a sudden
reduction in the lateral stiffness of the frame. Both from the monotonic curve and from the
displacement profiles it can be noted that the mechanism starts to develop at a value of top drift
close to 1%. The interstorey drift profile graph confirms how after the mechanism starts, a rapid
increase in floor deformation is recorded, reaching values that could not be sustained by the
structure.

- 117 -
Chapter 6 – Numerical investigation on 2D frame systems

Partially infilled frame


Top Displacement [m]
0 0.1 0.2
400

300
Base Shear [kN]

200

Panrtially Infilled (2 panel)


100
Partially Infilled (1 panel)

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.37: Pushover analysis on the 6-storey partially infilled frame: Base shear - Top drift curve

Top Drift [%]


0 0.4 0.8 1.2

6 6
16 16
Height [m]

Height [m]
4 12 4 12
Storey

Storey

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 2 4 6 8 10
Top Displacement [m] Interstorey Drift [%]

Figure 6.38: Pushover analysis on the 6-storey partially infilled frame: Displacements and Interstorey drift
profiles corresponding to top drift value of 0.5% and 1%.

The partially infilled frame, as already noted in the case of the 3-storey frame, shows a very similar
behavior regardless of the type of masonry infill adopted. Is believed that this type of response is
due the marked difference in strength and stiffness of the ground floor respect to the upper storey.
The base shear – top drift curve of the two infills solutions, when compared, show once again the
similarity in the seismic response of the frame. The effect of the reduced thickness of the masonry
panels is traduced in a slight difference in terms of elastic stiffness of the building.

- 118 -
Chapter 6 – Numerical investigation on 2D frame systems

Uniformly infilled frame


Top Displacement [m]
0 0.1 0.2
800

600
Base Shear [kN]

400

200
Uniformly Infilled (2 panel)
Uniformly Infilled (1 panel)

0
0 0.5 1 1.5
Top Drif t [%]
Figure 6.39: Pushover analysis on the 6-storey uniformly infilled frame: Base shear - Top drift curve

Top Drift [%]


0 0.4 0.8 1.2

6 6
16 16

Height [m]
4 12
Height [m]

4 12
Storey
Storey

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 4 8 12
Top Displacement [m] Interstorey Drift [%]

Figure 6.40: Pushover analysis on the 6-storey uniformly infilled frame: Displacements and Interstorey drift
profiles corresponding to top drift value of 0.5% and 1%.

The pushover analysis results of the 6-storey frames infilled with strong (solid line) and weak
(dashed line) infills are compared in the previous figures. The global behavior in terms of
displacement and drifts is very similar between the two solutions. In both cases a soft storey
mechanism develops at the 2nd storey as a consequence of the cracking and subsequent failure of the
infills panels. The lost of stiffness provided by the masonry infills leads to a high deformation
demand to the structural elements which rapidly yield and cause the development of the storey
mechanism.

- 119 -
Chapter 6 – Numerical investigation on 2D frame systems

On the other hand, the difference in strength and stiffness due to the presence of weak panels
instead of the strong ones is very evident looking at the base shear – top drift curve. The damaging
of the panels occurs almost at the same level of deformation, indicating that the damaging of the
infills panels is strictly linked to the deformation of the frame, but it happened at a value of force
almost 50% lower respect to the case of double panel infills.
Top Displacement [m]
0 0.1 0.2
800

Bare Frame
Uniformly Infilled Frame
600 Partial Infilled Frame
Base Shear [kN]

400

200

0
0 0.5 1 1.5
Top Drif t [%]

Figure 6.41: Comparison between the 6 storey frame pushover curves

In figure 6.41 a comparison between the base shear – top displacement curves of the analyzed
frames is reported. The graph highlights the influence of the presence of masonry infills on the
characteristics of the global structural response. The infill panels guarantee an increase of maximum
resistance and initial stiffness of the frame; after the failure of a group of panels and the consequent
development of a soft storey mechanism, the resistance of the structure decreases but still remaining
at an higher level respect to the bare frame, indicating that the presence of undamaged panels
contributes to the structural resistance even in the post-elastic range.

6.2.4 Time history analyses

The ten records described in paragraph 6.1.3 has been used to perform a series of time-history
analyses. In the following pages plots of the average results in terms of maximum displacement and
drift will be presented, for each of the structural solutions of the frame (bare, partially and

- 120 -
Chapter 6 – Numerical investigation on 2D frame systems

uniformly infilled). The same results will also be presented in tables together with the standard
deviation values, in order to give an immediate idea of the scatter that affects the analyses results.

Bare frame

6 6 18
5 5

Height [m]
4 4 12

Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

6 6 18
5 5

Height (m)
4 4 12
Storey
Storey

3 3

2 2 6
1 1

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Residual Interstorey Drift (%) Residual Displacement (m)

Figure 6.42: Time-history analysis on the 6-storey bare frame: interstorey drift and displacement profile
(average of maximum and residual)

Maximum displacement [m] Drift [%]


storey negative positive residual maximum residual
1 -0.022 (0.005) 0.022 (0.005) 0.0000 0.90 (0.16) 0.02
2 -0.081 (0.034) 0.100 (0.049) 0.0068 2.91 (1.51) 0.29
3 -0.116 (0.043) 0.141 (0.060) 0.0077 1.57 (0.38) 0.07
4 -0.202 (0.086) 0.225 (0.145) 0.0040 4.45 (3.21) 0.31
5 -0.251 (0.101) 0.278 (0.153) 0.0058 2.91 (0.71) 0.26
6 -0.275 (0.098) 0.307 (0.148) 0.0081 1.52 (0.34) 0.14
Table 6.10: Displacement and interstorey drift: mean values and standard deviation of ten analyses on the bare
frame

The analyses performed on the bare frame confirmed what was anticipated by the pushover
analysis. A soft storey mechanism is expected to develop at the 4th storey of the building where
- 121 -
Chapter 6 – Numerical investigation on 2D frame systems

reduction of columns cross sections is located. The maximum interstorey drift registered during the
ten analyses results in an average value of 4.45% (± 3.21%). The relative high value of standard
deviation indicates that the damage and consequent deformation is strictly correlated to the type of
record used in the analysis. The plots of residual displacement and interstorey drift show that some
permanent inelastic deformation has been experienced by structural elements located at the 4th floor
as a consequence of the failure of masonry infills. However the limited values of residual
displacement obtained can be explained considering that the model adopted in the analyses do not
take into account the member strength degradation and P-∆ effects.
In the following pages some more detailed results of the analyses performed using three particular
records (Loma Prieta (EQ4); Northridge (EQ8); Superstition Hills (EQ10)) are presented and the
maximum recorded values of beam-column joints rotation are reported.

- 122 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


Top Floor Displacement [m] 0.3 1.5

0.2 1

Top Drif t [%]


0.1 0.5
0 0
-0.1 -0.5
-0.2 -1
-0.3 -1.5

200
150
100
Base Shear [kN]

50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

6 6 18
5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18
5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.5 -0.3 -0.1 0.1 0.3 0.5
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.43: Bare frame: time history analysis results (Loma Prieta, EQ4)

- 123 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
Top Floor Displacement [m] 0.3 1.5

0.2 1

Top Drif t [%]


0.1 0.5
0 0
-0.1 -0.5
-0.2 -1
-0.3 -1.5

200
150
100
Base Shear [kN]

50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40
Time [sec]

6 6 18
5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18
5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.5 -0.3 -0.1 0.1 0.3 0.5
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.44: Bare frame: time history analysis results (Northridge, EQ8)

- 124 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills (EQ10)


Top Floor Displacement [m] 0.3 1.5

0.2 1

Top Drif t [%]


0.1 0.5
0 0
-0.1 -0.5
-0.2 -1
-0.3 -1.5

200
150
100
Base Shear [kN]

50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40
Time [sec]

6 6 18
5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18
5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.5 -0.3 -0.1 0.1 0.3 0.5
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.45: Bare frame: time history analysis results (Superstition Hills, EQ10)

- 125 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


6 6 6

4 4 4
Storey N°

Storey N°

Storey N°
2 2 2

0 0 0
0 0.004 0.008 0 0.004 0.008 0 0.004 0.008
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]

Column 1
Column 2
Column 3
Column 4

Figure 6.46: Bare frame joint rotations

Figure 6.46 shows the maximum values of joint rotation recorded during the three time history. The
joints that experience the higher deformation demands are the exterior ones. This confirms what has
already been noticed during the experimental tests and numerical analyses on the three storey R.C.
frame and on the beam-column subassemblies.

- 126 -
Chapter 6 – Numerical investigation on 2D frame systems

Partially infilled frame

6 6 18
5 5

Height [m]
4 4 12
Storey

Storey
3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

6 6 18
5 5

Height (m)
4 4 12
Storey

Storey

3 3
2 2 6
1 1
0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Residual Interstorey Drift (%) Residual Displacement (m)

Double panel
Single panel

Figure 6.47: Time-history analysis on the 6-storey partially infilled frame: interstorey drift and displacement
profile (average of maximum and residual)

Maximum displacement [m]


Double panel infill Single panel infill
storey Negative positive residual negative positive residual
1 -0.155 (0.056) 0.118 (0.056) 0.0012 -0.111 (0.059) 0.114 (0.061) 0.0008
2 -0.117 (0.057) 0.121 (0.056) 0.0012 -0.118 (0.061) 0.122 (0.064) 0.0008
3 -0.119 (0.057) 0.122 (0.056) 0.0012 -0.123 (0.061) 0.126 (0.064) 0.0008
4 -0.120 (0.057) 0.124 (0.056) 0.0012 -0.126 (0.062) 0.130 (0.064) 0.0008
5 -0.121 (0.057) 0.125 (0.056) 0.0012 -0.129 (0.062) 0.133 (0.064) 0.0008
6 -0.122 (0.057) 0.126 (0.056) 0.0013 -0.130 (0.062) 0.135 (0.064) 0.0008
Table 6.11: Displacement: mean values and standard deviation of ten analyses on the partially infilled frame

- 127 -
Chapter 6 – Numerical investigation on 2D frame systems

Drift [%]
Double panel infill Single panel infill
storey maximum residual maximum residual
1 5.23 (2.07 0.530 5.09 (2.29) 0.495
2 0.09 (0.01) 0.003 0.30 (0.09) 0.011
3 0.07 (0.00) 0.001 0.18 (0.01) 0.000
4 0.06 (0.00) 0.001 0.16 (0.02) 0.000
5 0.04 (0.00) 0.001 0.13 (0.02) 0.000
6 0.03 (0.00) 0.001 0.09 (0.01) 0.001
Table 6.12: Interstorey drift: mean values and standard deviation of ten analyses on the partially infilled frame

The response of the partially infilled frame system is governed by a soft storey mechanism located
at the 1st floor (without infills). In this case the infill panels suffer few or any damage since the
deformation of the frame is completely concentrated in the ground floor ad is sustained by the
column members. The type of behavior observed is consistent with what has been recorded during
the pushover analyses. A similar response has also been noticed during the analysis of the 3-storey
partially infilled frame, indicating that this structural solution is not very sensitive to the geometry
and dimension of the structure and that the difference in the building height doesn't influence the
global response of the structure. Also the type of record used to analyze the structure doesn't have a
big influence on the response of the frame, as indicated by the lower value of standard deviation of
displacement and drift, if compared to the other two solutions (bare and uniformly infilled).
In the following pages the results of the analyses using Loma Prieta (EQ4), Northridge (EQ8) and
Superstition Hills (EQ3) as input records are presented. In the following graphs the solid line
represent the response of the frame infilled with strong panels while the dashed line is
representative of the single panel infilled frame.

- 128 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


0.12 1
Top Floor Displacement [m]

0.08
0.5

Top Drif t [%]


0.04

0 0

-0.04
-0.5
-0.08

-0.12 -1
350
250
Base Shear [kN]

150
50
-50
-150
-250
-350
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

6 18
6
5 5

Height (m)
4 12
Storey

4
Storey

3 3
2 6
2
1 1

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05

Residual Interstorey Drift (%) Residual Displacement (m)

6 18
6
5 5
Height [m]

4 12
Storey

4
Storey

3 3

2 2 6

1 1

0 0 0
0 2 4 6 -0.2 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.48: Partially infilled frame: time history analysis results (Loma Prieta, EQ4)

- 129 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
0.16 1
Top Floor Displacement [m]

0.12
0.08 0.5

Top Drif t [%]


0.04
0 0
-0.04
-0.08 -0.5
-0.12
-0.16 -1
350
250
Base Shear [kN]

150
50
-50
-150
-250
-350
0 5 10 15 20 25 30 35 40
Time [sec]

6 18
6
5 5

Height (m)
4 12
Storey

4
Storey

3 3
2 6
2
1 1

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05

Residual Interstorey Drift (%) Residual Displacement (m)

6 18
6
5 5
Height [m]

4 12
Storey

4
Storey

3 3

2 2 6

1 1

0 0 0
0 2 4 6 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.49: Partially infilled frame: time history analysis results (Northridge, EQ8)

- 130 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills (EQ10)


0.16 1
Top Floor Displacement [m]

0.12
0.08 0.5

Top Drif t [%]


0.04
0 0
-0.04
-0.08 -0.5
-0.12
-0.16 -1
350
250
Base Shear [kN]

150
50
-50
-150
-250
-350
0 5 10 15 20 25 30 35 40
Time [sec]

6 18
6
5 5

Height (m)
4 12
Storey

4
Storey

3 3
2 6
2
1 1

0 0 0
0 0.4 0.8 1.2 -0.05 0 0.05

Residual Interstorey Drift (%) Residual Displacement (m)

6 18
6
5 5
Height [m]

4 12
Storey

4
Storey

3 3

2 2 6

1 1

0 0 0
0 2 4 6 -0.1 0 0.1 0.2
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.50: Partially infilled frame: time history analysis results (Superstition Hills, EQ10)

- 131 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


6 6 6

4 4 4

Storey N°
Storey N°
Storey N°

2 2 2

0 0 0
0 0.004 0.008 0 0.004 0.008 0 0.004 0.008
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]
a) b) c)
Column 1
Column 2
Column 3
Column 4

Figure 6.51: Partially infilled frame joint rotations

As expected, for the case of partially infilled frame, the maximum values of joint rotations were
recorded at the first storey. Again the exterior joints (Columns 1 and 4) are those which undergo the
higher level of inelastic deformation; thus indicating that cracking and damage of the panel region is
more likely to affect this typology of beam to column connection. Generally the deformation level
of the joints is not affected by the type of masonry infills adopted.

- 132 -
Chapter 6 – Numerical investigation on 2D frame systems

Uniformly infilled frame

6 6 18
5 5

Height [m]
4 4 12
Storey

Storey
3 3
2 2 6
1 1
0 0 0
0 0.4 0.8 1.2 -0.08 -0.04 0 0.04 0.08
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

6 6 18
5 5

Height (m)
4 4 12
Storey

Storey

3 3
2 2 6
1 1
0 0 0
0 0.2 0.4 0.6 0.8 1 -0.03 -0.015 0 0.015 0.03
Residual Interstorey Drift (%) Residual Displacement (m)

Double panel
Single panel

Figure 6.52: Time-history analysis on the 6-storey uniformly infilled frame: interstorey drift and displacement
profile (average of maximum and residual)

Maximum displacement [m]


Double panel infill Single panel infill
storey Negative positive residual negative positive residual
1 -0.013 (0.004) 0.016 (0.005) 0.0011 -0.014 (0.006) 0.016 (0.004) 0.0008
2 -0.033 (0.012) 0.047 (0.018) 0.0039 -0.039 (0.020) 0.049 (0.020) 0.0032
3 -0.043 (0.014) 0.057 (0.019) 0.0042 -0.048 (0.022) 0.060 (0.021) 0.0036
4 -0.050 (0.015) 0.064 (0.018) 0.0045 -0.055 (0.022) 0.068 (0.022) 0.0041
5 -0.055 (0.015) 0.070 (0.018) 0.0049 -0.060 (0.022) 0.074 (0.022) 0.0045
6 -0.059 (0.016) 0.074 (0.018) 0.0051 -0.064 (0.022) 0.078 (0.021) 0.0048
Table 6.13: Displacement: mean values and standard deviation of ten analyses on the uniformly infilled frame

- 133 -
Chapter 6 – Numerical investigation on 2D frame systems

Drift [%]
Double panel infill Single panel infill
storey maximum residual maximum residual
1 0.62 (0.15) 0.058 0.65 (0.19) 0.073
2 1.07 (0.51) 0.094 1.16 (0.58) 0.103
3 0.40 (0.07) 0.021 0.43 (0.11) 0.023
4 0.31 (0.04) 0.013 0.33 (0.07) 0.016
5 0.23 (0.03) 0.013 0.24 (0.02) 0.016
6 0.18 (0.03) 0.013 0.18 (0.02) 0.014
Table 6.14: Interstorey drift: mean values and standard deviation of ten analyses on the uniformly infilled frame

The seismic response of the uniformly infilled frame considered in the analyses is characterized by
a storey mechanism located at the second storey. The brittle failure of the masonry infills during the
seismic event, in fact, cause a sudden decrease of strength and stiffness at a particular level of the
building. As a consequence of this event, the structural elements (mainly columns) are subjected to
a high level of inelastic deformation demand, reaching values of interstorey drift of 1.07% (strong
infills) and 1.16% (weak infills). The residual displacement and drift plots show that the soft storey
mechanism at the 2nd level causes some permanent deformation that is an indicator of the damage
that the structure could suffer.
In the following pages some more detailed results of the analyses performed on the uniform infilled
frame with both strong (double panel) and weak (single panel) infills are reported. The single panel
infilled frame is indicated by the dashed line and the double panel masonry is represented by the
solid line. The maximum values of joint rotations are also reported as a measure of the contribution
of joints to the global deformation of the building.

- 134 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4)


0.1
Top Floor Displacement [m] 0.3
0.06
0.15

Top Drif t [%]


0.02
0
-0.02
-0.15
-0.06
-0.3
-0.1

800
600
400
Base Shear [kN]

200
0
-200
-400
-600
-800
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

6 6 18

5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18

5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6

1 1
0 0 0
0 0.3 0.6 0.9 1.2 -0.1 -0.05 0 0.05 0.1
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]

Figure 6.53: Uniform infilled frame: time history analysis results (Loma Prieta, EQ4)

- 135 -
Chapter 6 – Numerical investigation on 2D frame systems

Northridge (EQ8)
0.1
Top Floor Displacement [m] 0.3
0.06
0.15

Top Drif t [%]


0.02
0
-0.02
-0.15
-0.06
-0.3
-0.1

800
600
400
Base Shear [kN]

200
0
-200
-400
-600
-800
0 5 10 15 20 25 30 35 40
Time [sec]

6 6 18

5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18

5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6

1 1
0 0 0
0 0.5 1 1.5 -0.1 -0.05 0 0.05 0.1
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]
Figure 6.54: Uniform infilled frame: time history analysis results (Northridge, EQ8)

- 136 -
Chapter 6 – Numerical investigation on 2D frame systems

Superstition Hills (EQ10)


0.1
Top Floor Displacement [m] 0.3
0.06
0.15

Top Drif t [%]


0.02
0
-0.02
-0.15
-0.06
-0.3
-0.1

800
600
400
Base Shear [kN]

200
0
-200
-400
-600
-800
0 5 10 15 20 25 30 35 40 45 50 55
Time [sec]

6 6 18

5 5

Height (m)
4 4 12
Storey
Storey

3 3
2 2 6
1 1
0 0 0
0 0.2 0.4 0.6 0.8 1 -0.05 0 0.05
Residual Interstorey Drift (%) Residual Displacement (m)

6 6 18

5 5
Height [m]

4 4 12
Storey
Storey

3 3
2 2 6

1 1
0 0 0
0 0.5 1 -0.1 -0.05 0 0.05 0.1
Maximum Interstorey Drif t [%] Envelope Max Floor Displacements [m]
Figure 6.55: Uniform infilled frame: time history analysis results (Superstition Hills, EQ10)

- 137 -
Chapter 6 – Numerical investigation on 2D frame systems

Loma Prieta (EQ4) Northridge (EQ8) Superstition Hills (EQ10)


6 6 6

4 4 4
Storey N°

Storey N°
Storey N°
2 2 2

0 0 0
0 0.004 0.008 0 0.002 0.004 0.006 0.008 0 0.004 0.008
Joint rotation [rad] Joint rotation [rad] Joint rotation [rad]
a) b) c)
Column 1
Column 2
Column 3
Column 4

Figure 6.56: Uniformly infilled frame joint rotations

Also in the case of uniformly infilled frame system the exterior joint suffer the higher level of
rotation (i.e. shear deformation). The beam-column joints seem not to be influenced by the different
typology of masonry adopted for the infill panels; the values of rotation recorded in the case of
weak panel infills (dashed line) and strong infills (solid line) are very close to each other as shown
in figure 6.56.

- 138 -
Chapter 7 – Numerical investigation on 3D frame systems

Equation Section (Next)

7. NUMERICAL INVESTIGATION ON 3D FRAME SYSTEMS

In this chapter the seismic response of existing three dimensional reinforced concrete frames has
been investigated. Particular interest was turn on the effect of the direction of the input ground
motion on the overall response of the structure. Another interesting aspect is the possible occurrence
of torsional effects in the response of the buildings, in particular this phenomenon is more likely to
occur in three dimensional infilled frame as a consequence of sudden panel failure that can generate
some sort of asymmetry in the structural stiffness and strength.
The three dimensional analyses has been performed on two different frame systems. The first
building is a 3 storey 2/3 scaled frame that has been constructed by coupling two plane frames
(identical to the model tested an the Department of Structural Mechanics of the University of Pavia
and analyzed in the previous chapters) with weak transversal beams as it has been typically done in
the Italian design practice between 1950 and 1970. The second structure is a six storey R.C. frame
constituted of 3 parallel plane frames (described in par. 6.2) linked through weaker beams in the
transversal direction.
The frame structures has been considered with different distribution of infills (bare, partially and
uniformly infilled) and with single panel infills placed in the interior frames to represent the weak
partition walls typically present in residential buildings and stronger double panel infills along the
perimeter of the frame.
The two models have been analyzed through 3D pushover analyses and non-linear time history
analyses varying the direction of the input record has been performed on the three storey frame
model.

- 139 -
Chapter 7 – Numerical investigation on 3D frame systems

7.1. DESCRIPTION OF THE FRAMES

7.1.1 Three storey frame

The three dimensional 2/3 scaled frame is constituted of two parallel 3 bays plane frames identical
to the model tested at the Department of Structural Mechanics of the University of Pavia and
described in the present work in chapter 3. The two parallel frames are placed at distance of 3 m
(correspondent to 4.5 m in a full scale building) and are connected by four transversal beams. The
exterior beam sections are 330x200 mm (depth x width) correspondent to a 500x300 mm section in
the full scale frame. The interior beam sections are 103 mm (depth) x 535 mm (width) (i.e. 200x800
in prototype building). This type of wide section beams were very common in Italian structures
especially in interior frames in order to avoid architectural interferences that deeper beam sections
could generate. Figure 7.1 shows a plan view of the frame and the properties of the link beam
sections.

A A B B B B A A

Section A-A

Section B-B

12φ8
4φ8

6φ8
2φ8

Figure 7.1: Three storey 3D frame: plan view and beam section properties

- 140 -
Chapter 7 – Numerical investigation on 3D frame systems

The infills layout has been chosen as representative of a large number of existing buildings of the
Mediterranean area. The double panel infills, placed along the perimeter of the frame, represents the
exterior walls that are often constituted of two layers of bricks with an air inter-space in between
with insulating purpose. The weaker infills, constituted by a single layer of masonry blocks, are
representative of interior partition walls and are placed inside the bays of interior frames. Figure 7.2
shows the typical infills layout.

double panel infills

single panel infills

Figure 7.2: Three storey 3D frame: infills layout

7.1.2 Six storey frame

The six storey 3D frame is constituted of three plane frames, 3 bays each, which characteristics
have been already described in paragraph 4.2. The parallel plane frames are placed at an inter-axis
of 4.5 m and are linked by means of transversal beams, located in correspondence of the exterior
columns, and of a mixed masonry-concrete floor diaphragm. The diaphragm is 220 mm thick with a
top concrete slab of 40 mm and small concrete beams at an inter-axis of 400 mm symmetrically
reinforced with 2φ12 bars at top and bottom level of section. The exterior perimeter beams are
400x300 (depth x width) (Fig 7.4).
In figure 7.5 the layout of masonry infills adopted for the six storey three dimensional frame is
shown. The exterior walls have been modeled with double panel infills while the interior bays have
been infilled with weak masonry panels. The small bay (2.0 m) of the central frame has been left
free of infill panels to represent a possible location for the stair well.

- 141 -
Chapter 7 – Numerical investigation on 3D frame systems

4,50
9,00

4,50

4,50 2,00 4,50

11,00

Figure 7.3: Six storey frame plan view

Trave di bordo
Transversal Beam

A staffe Φ6/20 cm B A A B staffe Φ6/20 cm A

A B A A B A

4,5 4,5
9
2Φ12

2Φ12
2Φ12

Sezione A-A
Section A-A Sezione B-B
Section B-B

4Φ12
4φ12 2Φ12
2φ12
400
400

4φ12
4Φ12
2Φ12
2φ12

300 300

Figure 7.4: Transversal beam mechanical and geometrical properties

- 142 -
Chapter 7 – Numerical investigation on 3D frame systems

Double panela infills


tamponature doppio
paramento
(strong)

no infills
maglia vuota tamponature
Single panelainfills
singolo
paramento
(weak)

Figure 7.5: Six storey frame: infills layout

7.2. DESCRIPTION OF THE MODEL

As for the case of plane frames, the numerical model adopted for the analysis of the three
dimensional frames is based on the concentrated plasticity approach.
To adequately investigate the effective three dimensional response of the building, it was important
to describe the behavior of columns when subjected to bi-axial flexure together with axial load
variation. For this reason a three dimensional interaction surface has been assigned to column
members using an elliptic domain to represent the Mz – My interaction function.

Figure 7.6: Three dimensional interaction surface adopted for columns (Carr, 2004)

- 143 -
Chapter 7 – Numerical investigation on 3D frame systems

Similarly also beam-column joints three dimensional behavior is influenced by the interaction of bi-
axial shear deformation and axial load variation. In the model adopted an cyclic non-linear moment
– rotation relationship (par. 5.2.2 and 5.2.3) has been assigned to joint spring elements to describe
the effect of shear deformation of beam-column connections on the global response of the frame.
The bi-axial interaction in joints response has been neglected mainly because of a lack of
information available in literature regarding possible descriptions of the three dimensional
interaction surface for joint regions.

7.3. PUSHOVER ANALYSIS

The three dimensional frames has been preliminarily analyzed through pushover analyses
performed applying a triangular system of forces along the 2 principal directions of the buildings (X
and Z directions). The results of these analyses have been then compared with the response of the
frame obtained applying the load diagonally (45° respect to the principal X and Z axis). The
external force distribution along the height of the frame has been defined through the equation 6.3,
resulting in a triangular force pattern with the force modulus proportional to the floor mass and
height.

7.3.1 Three storey frame

Bare frame

Top Displacement [m]


0 0.03 0.06 0.09
140

120

100 bare frame


Base Shear [kN]

80

60

40
3D-X component
20 X pushover

0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.7: 3D bare frame pushover: X direction

- 144 -
Chapter 7 – Numerical investigation on 3D frame systems

Top Displacement [m]


0 0.03 0.06 0.09
140

120

100
Base Shear [kN]

80

60

40
3D-Z component
20 Z pushover

0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.8: 3D bare frame pushover: Z direction

Figures 7.7 and 7.8 show the comparison between the base shear – top drift curve obtained from the
pushover analysis on the three dimensional three storey bare frame. The solid line represents the
response of the frame in each of the principal directions when the external force are applied
diagonally (with an inclination of 45° respect to the principal axes). The dashed line, instead, is
obtained applying to the structure a system of force parallel to the X and Z direction respectively.
It appear evident that the resistance of the building decreases markedly when the frame is subjected
to a three dimensional force distribution. The structural element are subjected to a bi-axial
excitation that limits their resistance capacity.
This kind of response has been recorded both in the frame stronger direction (X) and in the weaker
one (Z).
Looking at the displacement profiles, shown in figure 7.9, it is worth noting how the mechanism
noted during the analysis of the 3 storey plane frame, is evident again in the 3D response. The effect
of joint damaging causes a hybrid soft storey mechanism that is no longer concentrated at one single
storey but involves the 1st and 2nd level of the structure. This kind of mechanism of shown in both
principal directions and it seems not to be influenced on the type of force system applied. The
results obtained confirm the important role played by the beam-column joint deformation on the
overall response of the structure.

- 145 -
Chapter 7 – Numerical investigation on 3D frame systems

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

3 6 3 6

Height [m]

Height [m]
Storey n°

Storey n°
2 4 2 4

1 2 1 2

0 0 0 0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08
Displacement [m] Displacement [m]

Figure 7.9: 3D bare frame pushover: displacement profiles

Partially infilled frame

Top Displacement [m]


0 0.03 0.06 0.09
180
160
140
Base Shear [kN]

120
100
80
60
3D-X component
40 X pushover
20
0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.10: 3D partially infilled frame pushover: X direction

- 146 -
Chapter 7 – Numerical investigation on 3D frame systems

Top Displacement [m]


0 0.03 0.06 0.09
180
160
140
Base Shear [kN]
120
100
80
60
40 3D-Z component
Z pushover
20
0
0 0.4 0.8 1.2
Top Drift [%]
Figure 7.11: 3D partially infilled frame pushover: Z direction

A behavior similar to the one shown by the bare frame has been recorded also for the partially
infilled frame. Again the resistance of the frame is higher when the applied force exert their action
along a single direction rather than in a three dimensional manner.
The response of the frame is governed by a soft storey mechanism located at the ground level (low
strength and stiffness), as it was expected and as the analyses on the plane frame anticipated.

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

3 6 3 6
Height [m]

Height [m]
Storey n°

Storey n°

2 4 2 4

1 2 1 2

0 0 0 0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08
Displacement [m] Displacement [m]

Figure 7.12: 3D partially infilled frame pushover: displacement profiles

- 147 -
Chapter 7 – Numerical investigation on 3D frame systems

Uniformly infilled frame

Top Displacement [m]


0 0.03 0.06 0.09
500
450
400
350
Base Shear [kN]

300
250
200
150
100
50
0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.13: 3D uniformly infilled frame pushover: X direction

Top Displacement [m]


0 0.03 0.06 0.09
500
450
400 3D-Z component
Z pushover
350
Base Shear [kN]

300
250
200
150
100
50
0
0 0.4 0.8 1.2
Top Drift [%]
Figure 7.14: 3D uniformly infilled frame pushover: Z direction

- 148 -
Chapter 7 – Numerical investigation on 3D frame systems

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

3 6 3 6

Height [m]

Height [m]
Storey n°

Storey n°
2 4 2 4

1 2 1 2

0 0 0 0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08
Displacement [m] Displacement [m]

Figure 7.15: 3D uniformly infilled frame pushover: displacement profiles

The response of the uniformly infilled frame is totally governed by the behavior of masonry infills,
the failure of which cause the development of a storey mechanism located at the 1st level of the
building. The three dimensional response reflects what has been anticipated by the analysis of the
plane frame model. Figure 7.15 shows the displacement profile recorded during the pushover
analysis for increasing level of top drift and it highlights how the sudden decrease of strength and
stiffness at the first storey, consequence of the brittle failure of masonry infills, leads to the soft
storey mechanism previously mentioned.
In this case the different direction in the application of external lateral force (principal direction vs.
45° degrees) doesn't influence markedly the global response of the frame as it can be noticed
comparing the base shear – top drift curves reported in figures 7.13 and 7.14.

- 149 -
Chapter 7 – Numerical investigation on 3D frame systems

7.3.2 Six storey frame

Bare Frame

The pushover analyses performed on the six storey 3D frame show how the behavior of the building
is characterized by a lower strength when it is subjected to a inclined force pattern (solid line) if
compared to the pushover curve obtained pushing the structure with a system of forces acting along
one of the principal axis only (dashed line). A similar behavior can be seen both along the strong
(X) and weak (Z) direction of the frame. This type of response confirms what has been obtained
from the analyses performed on the 3 storey frame.
Looking at the displacement profiles in the two principal directions (Figure 7.17 and 7.18) it is
worth noting that the displacement recorded in the case of 3 dimensional pushover appear
amplified, particularly evident in the lower storey, in compared to those obtained from a
unidirectional distribution of forces. A possible explanation can be found in the development of a
floor torsional deformation mechanism which causes an amplification of deformation at some
particular levels of the building.

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
160

140

120
Base Shear [kN]

100

80

60

40 3D-X component
X pushover
20

0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.16: 3D bare frame pushover: X direction

- 150 -
Chapter 7 – Numerical investigation on 3D frame systems

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
160

140

120 3D-Z component


Z pushover
Base Shear [kN]
100

80

60

40

20

0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.17: 3D bare frame pushover: Z direction

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

6 6
16 16
Height [m]

12

Height [m]
4 4 12
Storey n°

Storey n°

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 0.1 0.2
Displacement [m] Displacement [m]

Pushover 3D
Pushover 2D

Figure 7.18: 3D bare frame pushover: displacement profiles

- 151 -
Chapter 7 – Numerical investigation on 3D frame systems

Partially Infilled Frame

The base shear – top drift curves obtained from the 3D pushover on the partially infilled three storey
frame (Fig. 7.19 and 7.20) show how the response of the structure along one of the principal axis is
influenced by the direction of application of the forces. When the structure is subjected to a inclined
force pattern the behavior of the frame is characterized by a reduced value of strength and stiffness.
A similar trend in the three dimensional response of the frame has been obtained analyzing the bare
frame.
The displacement profiles along the X and Z direction, shown in figure 7.21, indicate that the
response of the frame structure is governed by a soft storey mechanism located at the first storey
and that the distribution of displacement along the height of the building is not influenced by the
direction of the applied forces. Regardless of the method of application of load, the big
discontinuity in terms of strength and stiffness between the ground and the upper levels, due to the
presence of masonry infills above the first storey governs the response of the structure.

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
500
450
400
350
Base Shear [kN]

300
250
200
150
3D-X component
100
X pushover
50
0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.19: 3D partially infilled frame pushover: X direction

- 152 -
Chapter 7 – Numerical investigation on 3D frame systems

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
500
450
400
350
Base Shear [kN]

300
250
200
150
100 3D-Z component
Z pushover
50
0
0 0.4 0.8 1.2
Top Drift [%]

Figure 7.20: 3D partially infilled frame pushover: Z direction

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

6 6
16 16
Height [m]

12

Height [m]
4 4 12
Storey n°

Storey n°

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 0.1 0.2
Displacement [m] Displacement [m]

Pushover 3D
Pushover 2D

Figure 7.21: 3D partially infilled frame pushover: displacement profiles

- 153 -
Chapter 7 – Numerical investigation on 3D frame systems

Uniformly Infilled Frame

The base shear – top drift curves of the uniformly infilled frame obtained from the pushover
analyses is characterized by a higher value of initial stiffness when the frame is subjected to a
inclined system of forces (solid line in the graphs of fig. 7.22 and 7.23). After the masonry infills
reach their maximum resistance the frame resistance suddenly decreases, showing a brittle behavior
even more evident if compared with the frame subjected to a unidirectional system of forces
(dashed line in the graph).
The displacement profile plots (figure 7.24) show that the response of the frame is governed by a
different mechanism depending on the type of load applied. The three dimensional pushover lead to
a story mechanism located at the ground floor where the infill panels reach first their maximum
resistance. When the structure is subjected to unidirectional loading, instead, the soft storey
mechanism develop at the second level.

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
2000
1800
3D-X component
1600
X pushover
1400
Base Shear [kN]

1200
1000
800
600
400
200
0
0 0.4 0.8 1.2
Top Drift [%]
Figure 7.22: 3D uniformly infilled frame pushover: X direction

- 154 -
Chapter 7 – Numerical investigation on 3D frame systems

Top Displacement [m]


0 0.05 0.1 0.15 0.2 0.25
2000
1800
3D-Z component
1600 Z pushover
1400
Base Shear [kN]

1200
1000
800
600
400
200
0
0 0.4 0.8 1.2
Top Drift [%]
Figure 7.23: 3D uniformly infilled frame pushover: Z direction

X Displacement profile Z Displacement prof ile

Top Drift [%] Top Drift [%]


0 0.5 1 1.5 0 0.5 1 1.5

6 6
16 16
Height [m]

12

Height [m]
4 4 12
Storey n°

Storey n°

8 8
2 2
4 4

0 0 0 0
0 0.1 0.2 0 0.1 0.2
Displacement [m] Displacement [m]

Pushover 3D
Pushover 2D

Figure 7.24: 3D uniformly infilled frame pushover: displacement profiles

- 155 -
Chapter 7 – Numerical investigation on 3D frame systems

7.4. TIME HISTORY ANALYSIS

The seismic response of the three dimensional three storey frame has been investigated through a
series of non-linear time history dynamic analyses using as input the ground motion set described in
par. 6.1.3. Each of the ten accelerograms used in the analyses has been applied to the structure in
four different directions, varying the inclination of the ground motion direction respect to the
principal X axis of the model. In particular the ground motions have been applied at 0° (i.e. parallel
to the X axis of the frame), 30°, 60° and 90° (parallel to the Z axis). Particular interest has been
given to the effect of the different angle of application of the input ground motion on the 3D
response of the building. In figure 7.25 the reference system for the frame model is shown.

Figure 7.25: Three storey frame model and reference system

- 156 -
Chapter 7 – Numerical investigation on 3D frame systems

Bare frame

The results of the time history analyses performed on the 3D three storey frame model are
summarized and presented in the following pages. Figure 7.26 shows, for each inclination of the
input ground motion, the average of the maximum storey displacement recorded analyzing the
structure with the 10 accelerograms considered. It is immediate to notice that when the ground
motion direction coincides with one of the principal direction of the structure (0° and 90°) the
influence of torsional effects on the structural response is negligible. On the other hand it is worth
noting that when the structure is subjected to an inclined input, the component of the maximum
displacement along the X and Z axis are comparable to those experienced during the analysis with
the accelerograms acting along the principal axis.

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.15 -0.05 0.05 0.15 -0.15 -0.05 0.05 0.15
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6
Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.15 -0.05 0.05 0.15 -0.15 -0.05 0.05 0.15
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

X direction
Z direction

Figure 7.26: Bare frame: maximum displacements

- 157 -
Chapter 7 – Numerical investigation on 3D frame systems

The maximum interstorey drift profiles (Fig. 7.27) indicate that, regardless of the attaching direction
of the input ground motion, the higher value of drift is experienced along the weak direction (Z
axis). This indicates that when the frame is subjected to an inclined accelerogram the torsional
response plays an important role in the overall behavior of the structure, amplifying the deformation
along the weaker direction of the building.

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

X direction
Z direction

Figure 7.27: Bare frame: maximum interstorey drift

- 158 -
Chapter 7 – Numerical investigation on 3D frame systems

Figures 7.28 and 7.29 show the profiles of average residual displacement and drift that give an
indication of the level of damage suffered by the frame. In particular looking at the residual drift
graphs it appears that the higher level on permanent deformation is reached at the first storey level,
confirming what has already been observed analyzing the 2D frame model.

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.015 -0.01 -0.005 0 0 0.005 0.01 0.015
Residual Displacement [m] Residual Displacement [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.015 -0.005 0.005 0.015 0 0.005 0.01 0.015
Residual Displacement [m] Residual Displacement [m]

X direction
Z direction

Figure 7.28: Bare frame: residual displacements

- 159 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

X direction
Z direction

Figure 7.29: Bare frame: residual interstorey drift

- 160 -
Chapter 7 – Numerical investigation on 3D frame systems

Partially infilled frame

The analyses performed on the partially infilled frame show that the response of the structure, with
this particular layout, is not influenced by the direction of the input ground motion. As expected the
behavior of the structure is governed by a storey mechanism located at the ground level. It is worth
noting how the maximum displacement and drift values, recorded during the analyses in the X and
Z directions, are comparable regardless of the direction of the input motion, thus indicating that the
big discontinuity in strength and stiffness between the first and second floor of the structure plays a
main role in influencing the response of the frame. Figure 7.30 and 7.31 show the plot of maximum
storey displacement and interstorey drift along the building height, averaged on ten time history
analyses.

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.15 -0.05 0.05 0.15 -0.15 -0.05 0.05 0.15
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6
Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.15 -0.05 0.05 0.15 -0.15 -0.05 0.05 0.15
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

X direction
Z direction

Figure 7.30: Partially infilled frame: maximum displacements

- 161 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 4 5 0 1 2 3 4 5
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 4 5 0 1 2 3 4 5
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

X direction
Z direction

Figure 7.31: Partially infilled frame: maximum interstorey drift

Figure 7.32 and 7.33 show the residual displacement and drift for the partially infilled frame. Again
the values shown seem not to be influenced by the input earthquake direction, confirming what has
been highlighted in the previous pages.

- 162 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.02 -0.015 -0.01 -0.005 0 -0.02 -0.01 0 0.01 0.02
Residual Displacement [m] Residual Displacement [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.02 -0.01 0 0.01 0.02 0 0.005 0.01 0.015 0.02
Residual Displacement [m] Residual Displacement [m]

X direction
Z direction

Figure 7.32: Partially infilled frame: residual displacements

- 163 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

X direction
Z direction

Figure 7.33: Partially infilled frame: residual interstorey drift

- 164 -
Chapter 7 – Numerical investigation on 3D frame systems

Uniformly infilled frame

Figure 7.34 and 7.35 show the average values of maximum storey displacement and drift obtained
during the time history analyses. Looking at the displacement and drift profiles it can be noticed that
when the structure is subjected to a input ground motion inclined at 60° respect to the X axis, the
response of the frame show a marked amplification in deformation. It is reasonable to suppose that
when the building is attacked along that particular direction, the sudden failure of a masonry infill
can cause an irregularity in the stiffness distribution, leading to a behavior governed by torsional
rotation that causes the amplification of storey deformation.

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.06 -0.02 0.02 0.06 -0.06 -0.02 0.02 0.06
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6
Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.06 -0.02 0.02 0.06 -0.06 -0.02 0.02 0.06
Envelope Max Floor Displacements [m] Envelope Max Floor Displacements [m]

X direction
Z direction

Figure 7.34: Uniformly infilled frame: maximum displacements

- 165 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Maximum Interstorey Drift [%] Maximum Interstorey Drift [%]

X direction
Z direction

Figure 7.35: Uniformly infilled frame: maximum interstorey drift

The graphs showing the profiles of residual displacement and drift (figure 7.36 and 7.37) confirm
the high level of deformation experienced by the first storey when the input ground motion attacks
the structure with a 60° inclination. The high value of residual displacement is also an indication of
an extensive level of damage suffered by the structural system and caused by the amplification in
displacement due to the torsional contribution.

- 166 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.05 -0.03 -0.01 -0.05 -0.03 -0.01
Residual Displacement [m] Residual Displacement [m]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
-0.05 -0.03 -0.01 -0.04 -0.02 0 0.02 0.04
Residual Displacement [m] Residual Displacement [m]

X direction
Z direction

Figure 7.36: Uniformly infilled frame: residual displacements

- 167 -
Chapter 7 – Numerical investigation on 3D frame systems

EQ Direction 0° EQ Direction 30°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

EQ Direction 60° EQ Direction 90°

3 6 3 6

Height [m]
Height [m]

Storey
Storey

2 4 2 4

1 2 1 2

0 0 0 0
0 1 2 3 0 1 2 3
Residual Interstorey Drift [%] Residual Interstorey Drift [%]

X direction
Z direction

Figure 7.37: Uniformly infilled frame: residual interstorey drift

- 168 -
Chapter 7 – Numerical investigation on 3D frame systems

Floor rotations

An interesting aspect to investigate is the amplitude of floor rotation that indicates the influence of
torsional effects on the overall structural behavior. Figure 7.37 show the average of the maximum
floor rotation obtained from the time history analyses performed on the 3D three storey frame.
The graphs show that the values of floor rotation are minimum when the input ground motion acts
along the frame principal axes, indicated as 0° (X axis) and 90° (Z axis) in the figures. As expected
the magnitude of the floor rotation reach it's maximum in the analyses performed considering the
earthquake attacking the frame with a 30° or 60° inclination respect to the principal X axis.
Generally the rotation of the floor diaphragm varies between 0 and 0.01 radiant. It is worth noting
that a very high value of rotation has been recorded from the analyses performed on the uniformly
infilled frame subjected to a 60° inclined ground motion. This confirms the type of behavior already
noted in the previous paragraph, indicating that the torsional deformation plays an important role on
the response of the uniformly infilled frame model. The seismic response of this particular structural
scheme can be influenced by the sudden failure of some masonry infill causing a situation of
stiffness and strength irregularity that is one of the main source of torsional deformation and
structural damage during a seismic event.

- 169 -
Chapter 7 – Numerical investigation on 3D frame systems

Bare frame Partialy infilled frame


0.03 0.03

Floor rotation [rad]


Floor rotation [rad]

0.02 0.02

0.01 0.01

0 0
0 30 60 90 120 0 30 60 90 120
Input inclination [°] Input inclination [°]

Unif ormly infilled f rame

0.03
Floor rotation [rad]

0.02

0.01

0
0 30 60 90 120
Input inclination [°]

Figure 7.38: 3D time history analysis: maximum floor diaphragm rotations

- 170 -
Chapter 8 – Conclusions

8. CONCLUSIONS

In the present work the assessment of the seismic response of pre-1970 reinforced concrete frames
with and without the presence of masonry infills has been investigated through non-linear numerical
analyses. Particular attention has been focused on the effect of mechanical properties variation and
geometrical distribution of infill panels on the global structural response. Two different type of
masonry panels have been considered: a strong double leaf panel representative of exterior walls
and a weaker single leaf panel typical of interior partition walls. This choice was also made with the
purpose of obtaining an upper and lower limit of results, thus giving a reasonable range for the
evaluation of seismic response of existing under-designed reinforced concrete frames.
From the results obtained in the present work it can be noted that the presence of masonry infills has
a dual effect on the overall structural response. On one hand the infill panels contribute to increase
the structural resistance against seismic action and their presence reduces the deformation demand
and consequent damage of structural elements. On the other hand an irregular distribution of panels
along building height can lead to the development of soft storey mechanism, as it has been
highlighted by the results obtained from the numerical analyses on the partially infilled frame
configuration. It is worth noting however that also when the infill panels are regularly distributed in
the frame (uniformly infilled frame), the seismic response of the structure was characterized by a
soft storey mechanism developing as a consequence of the brittle failure of masonry panels at a
particular level, that produces a sudden reduction of strength and stiffness and an increase in the
storey deformation demand.
The results of analyses performed on the infilled frame show that the typology of masonry panels,
whether single or double leaf, has a moderate influence on the global structural response. The
strength of the masonry infills gives a contribution to the maximum resistance of the building but,

- 171 -
Chapter 8 – Conclusions

considering the post-elastic response, the structural mechanism and the level of deformation
experienced by the frame are moderately influenced by the characteristics of the infill panels.
As it has been underlined in the previous chapters the structures designed between 1950 and 1970
for gravity loads only present a series of typical structural deficiencies among which the lack of
transverse reinforcement in beam-column joints and absence of any capacity design principles. For
these reasons the joint panel region often represents a weak link in the seismic structural response of
these type of buildings.
The results of the analyses performed in the present work confirmed the importance of modeling the
beam-column joints to better capture the global seismic response of the frame that can be strongly
influenced by the level of cracking and deformation experienced by the joint regions. The
experimental tests performed at the University of Pavia on a three storey R.C. frame and on beam-
column subassemblies (Calvi et al., 2002 and Pampanin et al., 2002), highlighted that the joints of
under-designed frame structures subjected to seismic loading experience extensive damage due to
shear deformation, combined with lack of transverse reinforcement and poor detailing. In particular
the higher damage observed was mainly concentrated in the exterior beam-column joints where, due
to the limited confinement provided by the surrounding structural elements, the high concentration
of stresses rapidly causes the formation of diagonal cracking of the joint panel region and an
increase of shear deformation demand. In the present work the beam column joints were modeled
using non-linear rotational spring with a constitutive law able to describe the loss of strength and
stiffness due to the formation of shear cracks and the "pinching" phenomenon consequence of the
slip of longitudinal reinforcing bars. The rotation of joint springs is strictly related to the shear
deformation demand and damage of the joint region. The results obtained from the non-linear
analyses performed show that the higher level of rotation is experienced by the springs located
along the exterior columns both for the 3-storey and the 6-storey frame building, thus confirming
the results of experimental tests. Furthermore it is worth noting that the same type of behavior is
also shown by the analyses on the partial and uniform infilled frames, suggesting that the different
distribution of infills panels in the frame doesn't modify this general behavior and the exterior joints
still result to be the more damaged.
The experimental tests performed by Calvi et al. (2002) highlighted that the cracking in the joint
and its consequent increase in deformation demand lead to the development of an hybrid failure
mechanism for the frame, in which the damaged joint acts as a "shear hinge", alternative to the
flexural plastic hinge, that contributes to spread the deformation demand on two adjacent storey.
The results obtained from pushover analyses performed on the three storey frame confirmed the
type of mechanism observed during the experimental tests, with the development of a soft storey

- 172 -
Chapter 8 – Conclusions

mechanism that interests the first and the second level of the frame. This particular type of structural
response, however, did not clearly emerged from the time history analyses performed on the 3
storey and 6 storey frames.
The results of time-history and pushover analyses presented in this work show that the structures
taken in consideration can sustain high levels of deflection with relatively small values of residual
deformations. In is to notice however that in reality beyond a certain level of deformation the
structure loses his capacity of bearing vertical loads, leading to the structural failure. This type of
mechanism wasn't captured because in the present work the P-∆ effects have not been considered in
the analyses.
The limited values of residual displacement that in some cases has been obtained from the analyses
can be explained considering that the model adopted doesn't take into consideration any strength
degradation relationship for structural elements and that no P-∆ effects have been included in the
analyses performed.
In the final chapter of the work the three dimensional response of pre-1970 reinforced concrete
frames has been investigated. The input seismic motion has been applied in four different directions
around the structure. Particular attention has been focused on the comparison between the two and
three-dimensional response of the frame and on the investigation of torsional effects, presented in
terms of floor rotations.

Further Investigations

For a better understanding of the three dimensional response of reinforced concrete frames an
extensive series of parametrical analyses is needed. In order to perform a high number of analyses a
reliable and refined 3D model is necessary, in particular the calibration of the biaxial interaction
surface for columns and the determination of a biaxial interaction relationship for joint elements is
important to correctly consider the influence of joint deformation on the structural response.
Another important feature is the consideration of P-D effects and strength degradation laws that can
significantly affect the post elastic response of the structure both at local and global level, giving a
more realistic description of the seismic behavior of existing R.C. frames.

- 173 -
Chapter 9 – References

9. REFERENCES

Baletta, G. "Risposta Sismica di Edifici a Telaio in C.A. con Tamponature Progettati per Soli
Carichi da Gravità", (in Italian) Laurea Thesis, Department of Structural Mechanics,
University of Pavia, 2002.
Bertoldi, S.H., Decanini, L.D. and Gavarini, C. "Telai tamponati soggetti ad azione sismica, un
modello semplificato: confronto sperimentale e numerico", (in Italian) atti del 6° convegno
nazionale ANIDIS, vol.2, pp.815-824, Perugia 13-15 Ottobre 1993.
Bing, L., Yiming, W. and Tso-Chien, P., (2003) “Seismic Behavior of Nonseismically Detailed
Interior Beam-Wides Column Joints-Part II: Theoretical Comparisons and Analytical
Studies” ACI Journal, V. 100, No. 1, January-February
Calvi, G.M., Magenes, G. and Pampanin, S. "Experimental Test on a Three Storey R.C. Frame
Designed for Gravity Only", 12th European Conference on Earthquake Engineering,
London, paper n. 727, 2002.
Carr, A.J. "Ruaumoko Program for Inelastic Dynamic Analysis – Users Manual", Department of
Civil Engineering, University of Canterbury, Christchurch, New Zealand, 2004.
Colangelo, F. "Experimental Evaluation of Member-by-Member Models and Damage Indices for
Infilled Frames", Journal of Earthquake Engineering, 7 (1), 25-50, 2003.
Colangelo, F. "Qualificazione, risposta sismica pseudodinamica e modelli fenomenologici di portali
di c.a. tamponati con laterizio", (in Italian) DISAT, Febbraio 1999.
Colombo, A., Negro, P. and Verzeletti, G. "Infilled frames: certainties and uncertainties", Proc. of
the 11th European Conference on Earthquake Engineering, CD-ROM, Paris, September 6-
11, 1998.

- 174 -
Chapter 9 – References

Cosenza, E., Manfredi, G. and Verderame, G.M. "Seismic Assessment of Gravity Load Designed
R.C. Frames: Critical Issues in Structural Modeling", Journal of Earthquake Engineering,
Vol. 6, Special Issue 1, 101-122, Imperial College Press, 2002.
Crisafulli, F.J. "Seismic Behaviour of Reinforced Concrete Structures with Masonry Infills", Ph.D.
Thesis, Department of Civil Engineering, University of Canterbury, July 1997.
Eligehausen, R., Ožbolt, J., Genesio,G., Hoehler, M. S and Pampanin, S. "Three-Dimensional
Modelling of Poorly Detailed Rc Frame Joints". Proceedings of the Annual NZSEE
Conference, Napier, March 2006.
Elmorsi M., Kianoush M. Reza. and Tso W.K. "Modeling bond-slip deformations in reinforced
concrete beam-column joints", Canadian Journal of Civil Engineering, 27: 490-505, 2000.
Eurocode 2 - Design of concrete structures - Part 1: General rules and rules for buildings, CEN
Technical Committee 250/SC2 1991., ENV 1992-1-1, Brussels.
Eurocode 8 - Design provisions for earthquake resistance of structures - Part 1-1: General rules -
Seismic actions and general requirements for structures, CEN Technical Committee
250/SC8 1994a., ENV 1998-1-1, Brussels.
Fardis, M. N. (editor), "Experimental and numerical investigations on the seismic response of rc
infilled frames and recommendations for code provisions", Report 6 of ECOEST-PREC8
Project, Laboratorio Nacional de Engenharia Civil, Lisbon, July 1997.
Fardis, M. N., Bousias, S. N. and Panagiotakos, T. B., "Seismic response and design of irregularly
infilled rc structures", Proc. of the 11th European Conference on Earthquake Engineering,
CD-ROM, Paris, September 6-11, 1998.
Fukada, Y. "Study on the Restoring Force Characteristics of Reinforce Concrete Buildings". (In
Japanese). Proc. Kanto District Symposium, Architectural Institute of Japan, Tokyo, Japan,
N°40, 1969
Galli, M. "Metodi per l'Analisi della Risposta Sismica di Edifici Esistenti con Struttura a Telaio in
Cemento Armato", (in Italian) Laurea Thesis, Department of Structural Mechanics,
University of Pavia, 2003.
Holmes, M. "Steel frames with brickwork and concrete infilling", Proceedings of the institution of
civil engineers, vol.19, 1961, pp.473-478.
Klingner, R.E. and Bertero, V.V. "Earthquake resistance of infilled frames", journal of the structural
division, vol.104, n.ST6, pp. 973-989, 1978.
Klingner, R.E., and Bertero, V.V. "Infilled frames in earthquake-resistant construction", report
EERC 76-32, Earthquake engineering research center, December 1976.

- 175 -
Chapter 9 – References

Leuchars, J.-M. "Masonry infills panels", E. report, University of Canterbury , February, 1973,
107pp.
Lowes, L.N. and Altoontash, A. "Modeling Reinforced-Concrete Beam-Column Joints Subjected to
Cyclic Loading", Journal of Structural Engineering, ASCE, pp. 1686 – 1697, December
2003.
Mainstone, R.J. "Supplementary note on the stiffness and strength of infilled frames", Current paper
CP13/74, Buildings research establishment, London 1974.
Monti, G. and Spacone, E. "Reinforced Concrete Fiber Beam Element with Bond-Slip", Journal of
Structural Engineering, pp. 654 – 661, June 2000.
Mosalam, K. M., White, R. N. and Gergely, P., "Seismic evaluation of frames with infill walls using
quasi-static experiments", Report NCEER-97-0019, State University of New York at
Buffalo, December 31, 1997.
Mosalam, K. M., White, R. N. and Gergely, P., "Seismic evaluation of frames with infill walls using
pseudo-dynamic experiments", Report NCEER-97-0020, State University of New York at
Buffalo, December 31, 1997.
Mosalam, K. M., White, R. N. and Gergely, P., "Computational strategies for frames with infill
walls: discrete and smeared crack analyses and seismic fragility", Report NCEER-97-0021,
State University of New York at Buffalo, December 31, 1997.
Nagai, T., Kashiwazaki, T. and Noguchi, H. “Three Dimensional Nonlinear Finite Element Analysis
of RC Interior Beam-Column Joints with Ultra High-Strength Materials Under Bi-
directional Load” Transactions of the Japan Concrete Institute, Vol. 18, 1996.
Negro, P., Anthoine, A., Combescure, D., Magonette, G., Molina, J., Pegon, P. and Verzeletti, G.
"Tests on Four-Storey Full-Scale Reinforced Concrete Frame with Masonry Infills:
Preliminary Report", Special publication No. I.95.54, European Commission, Joint Research
Centre, Ispra, Italy, 1995.
Otani, S. and Sake, A. "A Computer Program for Inelastic Response of R/C Frames to
Earthquakes". Report UILU-ENG-74-2029, Civil Engineering Studies, Univ. Of Illinois at
Urbana-Champaign, Nov. 1974
Pampanin, S., Magenes, G. and Carr, A.J. "Modeling of Shear Hinge Mechanism in Poorly Detailed
RC Beam/Column Joints", fib Symposium on Concrete Structures in Seismic Regions,
Athens, paper n. 171, 2002.
Pampanin, S., Moratti, M. and Calvi, G.M. "Seismic Behaviour of R.C. Beam-Column Joints
Designed for Gravity Loads", 12th European Conference on Earthquake Engineering,
London, paper n. 726, 2002.

- 176 -
Chapter 9 – References

Paulay, T. and Priestley, M.J.N. "Seismic design of reinforced concrete and masonry buildings",
John Wiley & Sons Inc., 1992, 744p.
Priestley, M.J.N., Seible, F. and Calvi, G.M. "Seismic Design and Retrofit of Bridges", John Wiley
and Sons, New York, 1996, 686pp.
Sacchi Mandriani, G. and Riccioni, R. "Comportamento statico e sismico delle strutture murarie"
(in Italian), CLUP Editore, Milano, 1982.
Stafford Smith, B. "Behaviour of square infilled frames", Proceedings of the American society of
civil engineering, journal of structural division, vol.92, n° ST1, 1996, pp.381-403.
Stafford Smith, B. and Carter, C. "A method of analysis for infilled frames", Proceedings of the
institution of civil engineers, vol.44, 1969, pp.31-48.
Takeda, T, Sozen, M. and Nielsen, N. "Reinforced Concrete Response to Simulated Earthquake",
ASCE Journal of Structural Division, 96(12), 1970.
Trowland, M. "Modelling the Shear Hinge in Beam Column Joints". Third Professional Year
Project, Department of Civil Engineering, University of Canterbury, New Zealand, 2003.
Youssef M. and Ghobarah A. "Modelling of RC beam column-joints and structural walls", Journal
of Earthquake Engineering, Vol. 5, No.1 (2001) 93-111.

- 177 -

You might also like