8/13/2019 Dissertation2003-Lupoi
1/78
Istituto Universitario
di Studi Superiori di PaviaUniversit degli Studi
di Pavia
EUROPEAN SCHOOL OF ADVANCED STUDIES IN
REDUCTION OF SEISMIC RISK
ROSE SCHOOL
LIMITATIONS AND PERFORMANCES
OF DIFFERENT APPROACHES FOR SEISMIC
ASSESSMENT
OF EXISTING BUILDINGS
A Dissertation Submitted in Partial
Fulfilment of the Requirements for the Master Degree in
EARTHQUAKE ENGINEERING
By
GIORGIO LUPOI
Supervisors: Prof. PAOLO EMILIO PINTO
Prof. GIAN MICHELE CALVI
June, 2003
8/13/2019 Dissertation2003-Lupoi
2/78
8/13/2019 Dissertation2003-Lupoi
3/78
The dissertation entitled Limitations and performances of different approaches for seismic
assessment of existing buildings, by Giorgio Lupoi, has been approved in partial fulfilment of
the requirements for the Master Degree in Earthquake Engineering.
Paolo Emilio Pinto ________________________________
Gian Michele Calvi ________________________________
8/13/2019 Dissertation2003-Lupoi
4/78
8/13/2019 Dissertation2003-Lupoi
5/78
ACKNOWLEDGEMENTS
The work has been carried out under partial funding from the EU project SPEAR (Contract No.
G6RD-CT-2001-00525).
8/13/2019 Dissertation2003-Lupoi
6/78
8/13/2019 Dissertation2003-Lupoi
7/78
ABSTRACT
The present study consists of a commented application of the three major guidance documents
on the assessment of existing buildings currently available: the New Zealand
Recommendations, the U.S. ASCE-FEMA356, and the Japanese Standard, to three structures
(two 2D and one 3D frames) which have been constructed at a large scale and tested. The main
purpose of the study is that of checking the practical applicability of the methods, the relative
ease of use, and of course the degree of agreement on the results.
The theoretical framework on which each document is based as well as the proposed methodsare outlined and commented. Differences of conceptual nature existing between the various
approaches are noted.
From the small number of cases examined is not possible to systematically trace the differences
in the results produced by the different approaches. The large difference in the way the shear
capacities of members and joints are evaluated has been a decisive factor in some cases for the
determination of the ultimate capacity of the entire building. However, even if this source of
discrepancy of the results from the various approaches was eliminated, the present exploration
indicates that significant differences would remain, linked to the criteria used to relate the
capacity curve to the response spectrum, or to the use of elastic analysis combined with local
ductility factors, as in the U.S. FEMA356, instead of the global mechanism analysis of New
Zealand.
8/13/2019 Dissertation2003-Lupoi
8/78
8/13/2019 Dissertation2003-Lupoi
9/78
v
TABLE OF CONTENTS
Aknowledgement ... i
Abstract ... iii
Table of contents ...... v
List of figures . vii
List of tables ix
1 INTRODUCTION.................................................................................................................. 1
2 REFERENCE DOCUMENTS .............................................................................................. 3
2.1 U.S. ASCE (FEMA 356) Prestandard ................................................................................. 3
2.1.1 Linear Static Procedure (LSP) ..................................................................... ............................5
2.1.2 Linear Dynamic Procedure (LDP) ................................................................... ........................6
2.1.3 Nonlinear Static Procedure (NSP) ........................................................... ................................6
2.1.4 Non-linear dynamic procedure (NDP) .............................................................. .......................7
2.1.5 Specific rules for concrete structures .................................................................... ...................8
2.2 New Zealand guidelines ...................................................................................................... 8
2.2.1 Force-based procedure ............................................................... ............................................ 10
2.2.2 Displacement-based procedure ....................................................... ....................................... 14
2.2.3 Comments ...................................................... ........................................................... .............15
2.3 Japanese Guidelines .......................................................................................................... 16
3 COMPARATIVE STUDY................................................................................................... 21
3.1 Selected procedures........................................................................................................... 21
3.2 The test structures ............................................................................................................. 21
3.2.1 Pavia Frame .............................................................. ............................................................. 21
3.2.2 Icons Frame............................................................................................................................23
3.2.3 Spear Frame ......................................................... ........................................................... .......24
3.3 Strengths of materials........................................................................................................ 25
3.4 Terms of comparison (capacity measure and earthquake input model)............................ 26
3.5 Non-linear numerical analysis........................................................................................... 27
8/13/2019 Dissertation2003-Lupoi
10/78
3.6 Risk reduction factor ......................................................................................................... 28
4 EXAMPLE APPLICATIONS............................................................................................. 29
4.1 FEMA 356: linear static .................................................................................................... 29
4.2 FEMA 356: Nonlinear static ............................................................................................. 32
4.3 NZ: Simple lateral Mechanism Analysis........................................................................... 34
4.4 NZ: Nonlinear Static.......................................................................................................... 38
4.5 BDPA: 3rdlevel ................................................................................................................. 41
5 COMPARISONS .................................................................................................................. 43
6 CONCLUSIONS...................................................................................................................51
7 REFERENCES ..................................................................................................................... 53
APPENDIX A: EC8 SPECTRA..................................................................................................57
APPENDIX B: MATERIALS MODELS ................................................................................. 59
A.1 Concrete............................................................................................................................. 59
A.2 Steel ...................................................................................................................................61
APPENDIX C: EARTHQUAKE INDUCED AXIAL FORCES.............................................. 63
8/13/2019 Dissertation2003-Lupoi
11/78
vii
LIST OF FIGURES
Figure 1. Mechanisms of post-elastic deformation of moment resisting frames [NZSEE 2002]11
Figure 2. Plastic displacement profiles for hinging at the base of the building [NZSEE 2002] . 13
Figure 3. Degradation of nominal shear stress of concrete for columns [Priestley 1996] .......... 14
Figure 4. Equivalent stiffness and equivalent damping for the Shibata-Sozen approach [NZSEE
2002] ................................................................................................................................... 15
Figure 5. View of Pavia frame ................................................................................................... 22
Figure 6. Gravity loads for Pavia frame...................................................................................... 22Figure 7. Reinforcement of columns and beams for Pavia frame............................................... 23
Figure 8. Four-storey frame, Icons.............................................................................................. 24
Figure 9. Gravity loads for Icons frame...................................................................................... 24
Figure 10. Geometric characteristics of the three test-structures ................................................ 25
Figure 11. Plant of Ispra frame and details of the columns reinforcement ................................. 25
Figure 12. EC8 elastic and design spectra .................................................................................. 26
Figure 13. Fiber model................................................................................................................ 27
Figure 14. Numbering of joints and sections and values of the coefficients of the vertical
distribution of the lateral load ............................................................................................. 30
Figure 15. Uniform and triangular vertical distributions of lateral loads.................................... 32
Figure 16. Resultant SDOF from mechanism analysis ............................................................... 37
Figure 17. Force approach of SLM for the Pavia Frame ............................................................ 37
Figure 18. Push-over curves for two types of distributions ........................................................ 38
Figure 19. Shear and flexural failures for the triangular distribution.......................................... 39
Figure 20. Design and elastic spectra for PGA at failure ............................................................ 40
Figure 21. Failure PGA for the Pavia frame................................................................................ 43
Figure 22. Failure PGA for the Icons frame................................................................................ 44
Figure 23. Collapse points of push-over procedures for Pavia frame......................................... 44
Figure 24. Collapse points of push-over procedures for Icons frame ......................................... 45
Figure 25. Comparison of shear capacities of joints for the Pavia and Icons frames.................. 45
Figure 26. Comparison of shear capacities of beams and columns for the Pavia and Icons frames
............................................................................................................................................. 46
Figure 27. Equivalent SDOF....................................................................................................... 46
8/13/2019 Dissertation2003-Lupoi
12/78
Figure 28. PGA s for different ductility levels ............................................................................ 48
Figure 29. Resisting frames of SPEAR structure ........................................................................ 49
Figure 30: Failure PGA for the 3D Spear building......................................................................49
Figure 31: EC8 acceleration spectrum Type 1, soil C.................................................................58
Figure 32: EC8 displacement spectrum Type 1, soil C...............................................................58
Figure 33: Stress strain relationship for different level of lateral confinement........................... 60
Figure 34: Stress strain relationship of the concrete ................................................................... 61
Figure 35: Stress strain relationship of the steel..........................................................................61
8/13/2019 Dissertation2003-Lupoi
13/78
ix
LIST OF TABLES
Table 1. Classification of vertical elements ................................................................................ 18
Table 2. Value of displacement compatibility factor.................................................................. 18
Table 3. Ductility index .............................................................................................................. 18
Table 4. Materials strength of the test structures (MPa) ............................................................. 26
Table 5. Values of minimum multiplier for each limit states ................................................. 31
Table 6. Collapse point for the four push-over analyses............................................................. 33
Table 7. Sway Indexes ................................................................................................................ 35Table 8. Base Shear at yielding and collapse.............................................................................. 35
Table 9. Curvature ductility and shear demand for the four base-columns ................................ 36
Table 10. Collapse point for the four analyses............................................................................ 40
Table 11. Results of the 3rdlevel procedure for the three different floors .................................. 42
Table 12. Failure mechanism according to the different procedures ......................................... 48
8/13/2019 Dissertation2003-Lupoi
14/78
8/13/2019 Dissertation2003-Lupoi
15/78
1 INTRODUCTION
Earthquake engineering experts, public Authorities and general public alike concur on the idea
that the assessment of the seismic safety of the built environment is a matter of high priority.
Awareness of the problem has been accelerated by the disastrous effects observed in recent
seismic events, in terms of loss of lives as well as of immediate and long-term economic losses.
The evolution of the attitude has been, one might say, more rapid that the capacity of the
technical community to cope adequately with it. Not because the seriousness of the problem had
escaped to it, but because the core activity had to be directed towards the improvement and the
harmonisation of the codes for the design of the new structures, on one hand, and also because
of the intrinsic difficulty of dealing with the problem of existing structures with procedures atthe same time rigorous, general, and practically applicable. This latter difficulty is compounded
with the lack of experimental data and models for the behaviour and the capacity of non-
seismically detailed members. The situation is now improving, with the appearance of recent
guidance documents on assessment from New Zealand [NZSEE, 2002], U.S.A. [ASCE, 2002]
and Japan [JBDPA, 1977-90], while Europe is about to complete its new Eurocode (Eurocode 8,
Part 3) on the subject.
Though each of these documents has been thoroughly checked for internal consistency,
comparative applications to selected structures have not been performed to date. Yet these test
applications are deemed to be interesting for a number of reasons: the procedures depart in
different ways from the direct design approaches, they make use of different models for
assessing the capacity of existing members, they provide more or less explicit and stringent
guidance to the user, and of course they offer different levels of complexity. Furthermore, trying
to explain results obtained from different procedures provides an excellent opportunity for
looking into the inner mechanisms of the procedures, and from this to pinpoint the more
relevant or critical aspects.
8/13/2019 Dissertation2003-Lupoi
16/78
2 Chapter 1. Introduction
The present study consists of a commented application of the three documents mentioned above
to the three structures (two 2D and one 3D frames) which have been constructed at a large scale
and tested (except the 3D one at the time of this writing) in the experimental facilities of the
University of Pavia and of the JRC in Ispra. Complete knowledge of the material properties,
reinforcement details and of the results from the actual tests are the reasons behind this
particular choice.
8/13/2019 Dissertation2003-Lupoi
17/78
2 REFERENCE DOCUMENTS
This section provides a brief general introduction to the three regulatory documents used in this
study, in order to highlight their main features as well as the principal differences among them.
2.1 U.S. ASCE (FEMA 356) Prestandard
The ASCE Prestandard for the Seismic Rehabilitation of Buildings, commonly known as
FEMA 356, is a comprehensive State-of-the-Art document on rational quantitative seismic
assessment and rehabilitation of existing buildings of concrete, steel or cast iron, masonry and
timber construction, and represents the most recent US development in seismic assessment. In
broad terms, it can be said to belong to the category of the displacement based approaches.
A brief summary is provided hereafter, focusing on the aspects relevant to the assessment of
concrete structures.
The assessment process starts with the definition of one or more Rehabilitation Objectives, each
one consisting on the selection of both a Target Building Performance Level and of the
corresponding Seismic Hazard Level. Several options are available. The Building
Performance Level is given by a combination of a Structural Performance Level, varying
from Immediate Occupancy to Collapse Prevention, and a Non Structural PerformanceLevel, varying from Operational to Not-Considered.
The seismic hazard can be represented either by an acceleration response spectrum or by
acceleration time histories. Two basic Earthquake Hazard Levels are defined: Basic Safety
Earthquake 1, corresponding to an event with 10% probability of exceedance in 50 years (return
period equal to 475 years), and Basic Safety Earthquake 2, corresponding to an event with 2%
probability of exceedance in 50 years (return period equal to 2475 years). A response spectrum
for different soil types and damping coefficients is provided as function of two parameters: the
8/13/2019 Dissertation2003-Lupoi
18/78
4 Chapter 2. Reference Documents
short-period response acceleration, XSS , and the long-period response acceleration, 1XS , which
have to be determined using the values taken from approved maps developed by the United
States Geological Survey and modified for the soil class of the site.
The selection of the so-called knowledge factor comes next. It accounts for the
uncertainties in the collection of as-built data and depends on the selected Rehabilitation
Objective, the accuracy of the data collected and the selected analysis procedure. The values for
are tabled in the Prestandard, varying from 0.75 to 1.
The actual assessment of the building can now be undertaken, according to one of the two
possible Rehabilitation Methods: the Simplified Rehabilitation Method and the Systematic
Rehabilitation Method. The former may be applied to certain buildings of regular configuration
that do not require advanced analytical procedures; the latter may be applied to any building
and involves thorough checking of each existing structural element or component, . , and the
verification of acceptable overall performance represented by expected displacements and
internal forces.
In this study, attention is concentrated on the more accurate method, i.e. the Systematic
Rehabilitation. Four different analysis procedures are allowed for the evaluation of the response
of the building: the linear static (LSP), the linear dynamic (LDP), the non-linear static (NSP)
and the non-linear dynamic (NDP).
The two linear procedures (LSP and LDP) are permitted only for buildings with a regular
structural configuration. A regular configuration is defined by means of geometrical
requirements and of a parameter called demand-capacity ratio (DCR), which is defined as the
ratio between the force acting on a member due to gravity and earthquake loads and its expected
strength. This parameter is intended to be a rough evaluation of magnitude and distribution of
the inelastic demands in the building; it is used however only to determine the structural
regularity.
The static procedures (LSP and NSP) are permitted only for those structures in which higher
mode effects are not significant. No limitations are set to the use of the non-linear dynamicprocedure.
The members are classified as primary or secondary according to their function: primary
elements are those that provide the capacity of the structure to resist collapse under seismic
force, secondary elements are all the others. Only primary elements are included in the
mathematical model of the structure. Detailed instructions are provided to include in the
analysis torsional as well as P effects.
8/13/2019 Dissertation2003-Lupoi
19/78
U.S. ASCE (Fema 356) Prestandard 5
The internal actions on members are classified as either deformation-controlled or force
controlled: the first ones are characterised by a ductile behaviour, the second ones by a brittle
behaviour. The classification is not up to the user, but it is explicitly indicated in the document.
In RC frame-type buildings, for example, the bending moment in beams and columns is
considered as a deformation-controlled action, while the shear is considered as a force-
controlled action.
This distinction governs the way in which verification of the elements is carried out. In the first
place, it affects the evaluation of the elements capacity: the expected value for the materials
strength is used for deformation-controlled actions, a lower-bound estimate for force-controlled
actions. More importantly, the limit state equation that governs the capacity/demand comparison
changes in accordance with the type of action: this latter point will be illustrated in detail
separately for each of the analysis procedures.
2.1.1 Linear Static Procedure (LSP)
The building response is evaluated by a linear elastic analysis. The seismic action is modelled as
a horizontal lateral load V, whose intensity is determined by the following expression
WSCCCCV am321= , (1)
where: 1C is a modification factor relating the expected maximum inelastic displacements to
the calculated elastic response; 2C represents the effect of stiffness degradation and strength
deterioration on maximum displacement response; 3C accounts for the increase of the
displacements due to P effects; mC is the effective weight factor to account for higher
modal mass participation effects, depending on the structural typology and on the number of
storeys; aS is the Response Spectral Acceleration at the fundamental period and damping ratio
of the building in the direction under consideration and W is the total weight of the building.
The lateral load Vis vertically distributed at floor levels proportionally to the floor mass and to
the floor height.
The fundamental period of the building can be estimated using either an empirical expression
provided in the document, or by an eigenvalue analysis, or by the Rayleigh-Ritz method.
A capacity/demand approach is applied to carry out the assessment for each individual element.
For deformation-controlled actions, the limit-state equation reads:
EGCE QQQm . (2)
In the left hand side of Eq. (2), the member capacity is given by the product of the expected
strength CEQ times the knowledge factor times the local ductility factor m . The latter is
8/13/2019 Dissertation2003-Lupoi
20/78
6 Chapter 2. Reference Documents
introduced to account for the expected ductility associated with this action. Values of m are
given as functions of section geometry, amount of reinforcement, amount of the shear force and
selected performance level. For the Collapse Prevention performance level the values of m
for beams range between 7 and 2.
On the right hand side of Eq. (2) the demand on the members is the sum of the gravity load
effects GQ and of the earthquake load effects EQ .
For force-controlled actions, the limit-state equation reads:
JCCC
QQQ EGCL
321
. (3)
The member capacity is the product of the lower-bound strength CLQ times the knowledge
factor . The effects of the earthquake loads are reduced by the coefficients1
C ,2
C ,3
C
already introduced in Eq. (1) and by the force-reduction factor J , which is greater than or equal
to 1.0. Its value may be taken as the smallest DCR of the components or, alternatively, equal to
2.0, 1.5 and 1.0 in Zones of High, Moderate and Low Seismicity, respectively.
The limit-state equations reflect the failure mechanisms associated with the two types of
actions: in the case of a ductile behaviour, Eq. (2), the capacity is increased by the ductility
factor m ; in the case of a brittle behaviour, Eq. (3), the external elastic action is reduced by the
coefficient J , to account for the inelastic behaviour of the elements delivering load to the
brittle element.
Although the verifications for both force-controlled and deformation-controlled actions are
expressed in terms of forces, the procedure belongs to the category of the displacement-based
approaches, since its qualifying feature is the implicit adoption of the equal displacement rule
(adjusted by the factor 1C ).
2.1.2 Linear Dynamic Procedure (LDP)
The only difference with respect to the LSP is the method of analysis: a modal spectral analysis
is carried out using un-reduced linear elastic response spectra. The members verifications are
analogous to those of the LSP.
2.1.3 Nonlinear Static Procedure (NSP)
The demand on the building is calculated by means of a push-over analysis. The inelastic
behaviour of the structure is explicitly accounted for in the numerical model by use of either
concentrated or diffused plasticity element formulations. A monotonically increasing lateral
load is applied to the building until a target displacement t is reached at a control node, taken
8/13/2019 Dissertation2003-Lupoi
21/78
U.S. ASCE (Fema 356) Prestandard 7
to be the centre of mass at the roof of the building. The target displacement is intended to
represent the maximum displacement likely to be experienced during the design earthquake
and it is determined by the expression:
g4
TSCCCC
2
2e
a3210
=t .(4)
where 0C is a modification factor relating the spectral displacement of an equivalent single
degree of freedom (SDOF) system to the roof displacement of the building, and 1C , 2C , 3C and
aS have been defined in section 2.1.1.
The push-over analysis yields a relation between the base shear and the displacement at the
control node. This force/displacement curve is bi-linearised and the effective fundamental
period of the building calculated according to the expression:
e
iie
K
KTT = . (5)
where iT is the elastic fundamental period, iK is the elastic lateral stiffness, eK is the effective
lateral stiffness.
The format of the limit state equation is common to both displacement-controlled and force-
controlled actions:
( )tDC > . (6)
The Eq. (6) says that the member capacity Chas to be greater than the demand D evaluated at
the target displacement t . It is worth noting that for deformation-controlled action, the check is
made in terms of ultimate rotation capacity and demand, not in terms of bending moments, as in
the two linear procedures (LSP and LDP). The expected rotational capacity is tabled in the
document for several typical section and reinforcement configurations.
A further condition is put on the base shear at the target displacement, which cannot be less than
80% of the effective yield strength of the structure.
2.1.4 Non-linear dynamic procedure (NDP)
The numerical model of the building and the limit-state equations are similar to those of the
NSP. The only difference is that the structural response is calculated using a Time History
Analysis.
8/13/2019 Dissertation2003-Lupoi
22/78
8 Chapter 2. Reference Documents
2.1.5 Specific rules for concrete structures
The flexural strength and the deformation capacity of members are calculated in accordance
with either the procedures of ACI 318 or other approved methods. Reductions in the
deformation capacity due to the shear have to be taken into consideration.
The shear strength has to be calculated in accordance with the procedures of ACI 318. A
reduction of 50% of the contribution of the transverse reinforcement is required if the
longitudinal spacing of transverse reinforcements exceeds half of the component effective
depth.
In setting up the numerical model, the members stiffness has to be calculated taking into
account the effective state of stress as well as the bar-slippage phenomenon. The reduction
coefficients for the members initial stiffness, to be applied in the case of linear analysis, are
explicitly indicated in the document. For example, the effective stiffness of beams is taken as
equal to half of the initial, un-cracked, stiffness. In columns, the reduction depends on the stress
state under gravity load; it varies form 0.5 to 0.7 of the initial stiffness.
2.2 New Zealand guidelines
The document produced by the New Zealand Society for Earthquake Engineering provides a
means of assessing the capability of existing buildings to reach an adequate level of seismic
performance and, where found necessary, improving the seismic performance up to a certain
level. A first draft was circulated for comment in 1996 and a second general draft was
completed in 2002.
Two levels of assessment are considered in the document: an Initial Evaluation of the state of
the building and a Detailed Assessment. The Initial Evaluation is a coarse screening, similar
to the simplified procedure of FEMA 356, which is intended to provide an approximate
assessment of the likely performance of a building in an earthquake. The Detailed Assessment is
intended to evaluate with greater accuracy the capability of an existing building to perform
satisfactorily under an earthquake of selected level of intensity through application of more
advanced procedures. In this study the attention is concentrated on the Detailed Assessment.
The focus of the detailed procedure is the determination of demand on structural elements,
resulting form the response of the building, and the assessment of the capacity of such elements
to meet the demand without causing loss of structural integrity. The NZ document concentrates
only on matters relating to life safety, i.e. collapse which leads to loss of life. The ultimate limit
state (ULS) is selected as boundary between the acceptable and the unacceptable performance.
8/13/2019 Dissertation2003-Lupoi
23/78
New Zealand guidelines 9
The ULS is conventionally reached when the building or any part of it: (a) loses stability, (b)
exceeds prescribed displacement limits or (c) is strained to the accepted limits of the (structural)
materials involved.
Three possible approaches for performing the assessment are indicated in the document: time
history analysis, force analysis and displacement analysis. Leaving out the first, which is by far
the most accurate but the most complex as well, the attention is set on the other two approaches.
In the document it is stated that the displacement-based approach is generally considered to
produce more rational and less conservative assessment outcome, the force-based one is more
familiar to designers. This is as far as the document goes in assessing the relative merits of the
two approaches.
Four analysis methods of different level of sophistication are proposed for the evaluation of the
structural response. Two of them rely on linear elastic theory: a traditional equivalent static
analysis and a modal response spectrum analysis. Limitations to the use of the elastic methods
similar to those of FEMA are set in the NZ document. The other two consider inelastic
behaviour; these are: Simple Lateral Mechanism analysis (SLM) and lateral push-over
analysis. It is observed that the proposed procedures are the same as those in the FEMA
document, with the exception of the SLM that is a peculiarity of the NZ document. This
approach involves a hand analysis to determine the probable collapse mechanism, its lateral
strength and displacement capacity, with the help of simplified considerations of capacity issues
(relative strengths in flexure and shear, etc.). The behaviour of the structure is reduced to that of
an equivalent SDOF system.
The main weakness of SLM approach is the inability to identify the sequence of development of
inelastic action between different members of the structure, a result achievable by means of a
push-over simulation.
The model of the earthquake action depends on the analysis method applied. The acceleration
response spectra and the displacement response spectra to be used in the assessment according
to force-based or displacement-based approaches, respectively, are those defined in the NZ
loading code [SNZ, 1976].
A brief outline of both force-based and displacement-based procedures is given in the
following, in the same way as they are presented in the original document. As it will be seen,
the route followed to determine the likely lateral inelastic mechanism and its
strength/deformation capacity is common to both procedures, i.e. the SLM analysis method and
the push-over, the essential difference intervening in the way the structural response is
evaluated.
8/13/2019 Dissertation2003-Lupoi
24/78
10 Chapter 2. Reference Documents
2.2.1 Force-based procedure
The procedure starts with the evaluation of members capacities. The probable flexural strengths
are calculated according to standard theory. The shear strengths in beams, columns and joints
are derived by means of the expressions provided in the document. A strength reduction factor
equal to 0.85 is applied to account for the approximation in the shear theory. The calculations
are to be based on the expected values for the materials strengths. If the data available are not
reliable, a variation of either one standard deviation or +/- 20% of the mean is suggested. It is
required to consider the earthquake-induced axial force in the calculation of the flexural and
shear capacities of the columns.
The post-elastic critical mechanism is investigated next. To this end, thesway potential index is
introduced. It is defined as the ratio between the sums, for all the joints at that horizontal level,
of the probable flexural strengths of beams over the probable flexural strengths of columns:
( )
( )+
+=
cbca
brbl
iMM
MMS , (7)
where blM and brM are the flexural capacities of the beams at left and right of the joint and
caM and cbM refer to the columns above and below the joint.
If 85.0>iS , the NZ document suggests that plastic hinges would develop in the columns;
otherwise they would develop in the beams. The most likely post-elastic mechanism may be
established from the calculation of the iS s at all levels. The possible outcomes are: a column
sidesway mechanism, a beam sidesway mechanism or a mixed mechanism: see Figure 1. In the
latter case, column plastic hinges would occur at some levels and beam plastic hinges at others.
The possibility of members shear failures has not been yet investigated at this point of the
procedure.
If the mechanism of post-elastic deformation is obvious from the onset, the SLM analysis
method can be applied for the calculation of the probable lateral seismic force capacity, bV . For
example, if a column sidesway mechanism is detected at the bottom floor, the probable lateralforce capacity of the frame is given by the sum of the shear forces in the columns of that storey,
found from the sum of the probable flexural strengths of the plastic hinges at the top and the
bottom of the columns of that storey divided by the storey height. This estimate is an upper
bound of the lateral force capacity.
8/13/2019 Dissertation2003-Lupoi
25/78
New Zealand guidelines 11
Figure 1. Mechanisms of post-elastic deformation of moment resisting frames [NZSEE 2002]
If the sway-index approach does not provide a clear indication on the failure mechanism, either
a standard linear elastic analysis or a non-linear push-over analysis have to be carried out. In the
first case, a linear-elastic model of the frame is set up; the equivalent static horizontal forces are
increased until the first plastic hinge forms. This method provides a lower bound to the probable
lateral force capacity of the frame, not accounting for moment redistributions. In the second
case, the inelastic deformation capacity of members is explicitly taken into account in the
structural model and the applied lateral loads are increased until the flat branch of the
force/displacement curve is reached. The NZ document suggests to apply both an inverted
triangular and a uniform horizontal force distribution.
The push-over analysis is more accurate than the SLM or the LSP, but it is also more complex.
The benefit of using the SLM approach is obvious, since it does not require numerical
modelling of the structure.
The probable base shear coefficient, i.e. the spectral acceleration at failure, is calculated by
means of the expression:
( )
RZSW
TC
pt
bsdh
V=, , (8)
where tW is the total weight of the structure, pS is the structural performance factor, R is the
return period factor based on the life span of the structure, Zis the zone factor and corresponds
to the response acceleration at 0.5 sec on a rock site with a 500-year return period. The values of
Zand R are specified in the document; their product defines the anchoring of the spectrum.
The force-based approach comes here into play: using the appropriate seismic hazard
acceleration spectra of NZS 4203:1192, the required structural ductility factor sd is derived
on the basis of hC and of the elastic period of the frame T. The NZ document does not indicate
8/13/2019 Dissertation2003-Lupoi
26/78
12 Chapter 2. Reference Documents
explicitly how to compute the fundamental period of the structure; it is however required to
include the effects of cracking on the section properties. The use of the acceleration spectrum
implicitly assumes that the structure behaviour can be represented as that of a SDOF oscillator.
The assessment whether the required demand sd can be satisfied by the available structural
ductility sc comes next. Three methods for the evaluation of sc are outlined in the NZ
document. The first one is a simple and qualitative evaluation of sc , based on the most likely
mechanism and on the detailing of the lateral reinforcement. A second possibility is to derive
the global displacement ductility sc directly from the (local) inelastic capacity of members
sections. To this end, empirical expressions of the plastic deformation are provided in the
document for the cases of beam sidesway and column sidesway mechanisms (see Figure 2,
which refers to plastic hinging occurring at the base of the building):
mmhLh ePycepp 391 === , (9)
where p is the plastic rotation of the column, eh is the equivalent height of the building, c
is the curvature ductility, y is the yielding curvature and pL is the plastic-hinge length. The
latter is function of the diameter of the longitudinal reinforcement bd and the yield strength of
the steel yf :
ybP fdLL += 022.008.0 , (10)
with L being the length of the member.
In addition lateral displacement of the frame at first yield y has to be evaluated. In the
document it is generically suggested to carry out a linear elastic push-over analysis for the
determination of y (it is believed that the analysis has to stop when the yielding moment in
any of the members sections is reached). This solution, however, represents a limitation to the
use of the SLM method, whose key feature is to avoid numerical modelling. Alternatively, an
empirical expression for y is provided in the context of a development of the SLM approach
proposed by Priestley into an appendix of the document:
= e
b
byy h
h
l5.0 , (11)
where y is the steel yielding strain, bl is the length of the shortest span, eh is the equivalent
height of the building and bh is the beam section height.
8/13/2019 Dissertation2003-Lupoi
27/78
New Zealand guidelines 13
Figure 2. Plastic displacement profiles for hinging at the base of the building [NZSEE 2002]
The third method for the evaluation of sc is the non-linear static push-over: the lateral forces
are increased until the available ultimate curvature is reached at the critical plastic hinge. The
ultimate displacement is then obtained from numerical simulation. This option represents the
most advanced and accurate method available thus far.
If sdsc < , the procedure is terminated, with the conclusion that structure needs to be
retrofitted. If sdsc > , a check whether shear failures occur before the ultimate flexural plastic
rotations are reached needs to be carried out.
For each member, the reduction of the shear strength due to the curvature ductility demand
associated with the development of plastic hinges is estimated using the degradation laws
proposed by Priestley and co-workers (see for example Figure 3). The reduced shear strength,
crV , is compared with the shear demand on member generated by the plastic hinges (flexural)
mechanism, flV , the latter obtained from the equilibrium condition considering the ultimate
values of bending moments. If flcr VV < , strengthening is necessary.
A final check on the inter-storey drift is required to ensure against significant P effects and
damage to non-structural elements.
8/13/2019 Dissertation2003-Lupoi
28/78
14 Chapter 2. Reference Documents
Figure 3. Degradation of nominal shear stress of concrete for columns [Priestley 1996]
2.2.2 Displacement-based procedure
The first part of the procedure dealing with the calculation of the members strengths, theidentification of most-likely failure mechanism and the evaluation of the frame probable lateral
force capacity, is the same as the force-based procedure.
The next step consists in checking the actual available shear strength, accounting for its possible
reduction due to curvature ductility demand, following the same procedure as described in the
previous section. Following this check the available local rotation capacities are reduced, if
necessary, to the values pertaining to actual shear capacity. It is noted that this check on shear is
anticipated with respect to the force-based procedure.
The global ductility capacity of the structure, sc , is found as in the force-based procedure. The
available displacement capacity at failure is obtained simply as:
( )scysc += 1 . (12)
For the evaluation of the yielding displacement, the indications given in the previous section
apply.
The demand due to the earthquake action is now evaluated applying the substitute-structure
approach [Shibata and Sozen 1976]. The frame behaviour is again reduced to that of a SDOF
oscillator, but the effective stiffness at maximum displacement is used:
sc
beff
VK= , (13)
from which the effective period T for the structure is calculated as:
effK
MT 2= , (14)
where is the mass of the structure.
8/13/2019 Dissertation2003-Lupoi
29/78
New Zealand guidelines 15
The maximum displacement demand sd is found from a set of displacement response spectra,
relative to different levels of the equivalent viscous damping, v . The level of damping of the
frame depends on the ductility capacity sc and on the predominant form of plastic hinging
developed. Curves are provided relating v to sc for the most common failure mechanisms
(see Figure 4).
Thus, in the displacement procedure the seismic response is characterised by an effective
stiffness and an equivalent damping, rather than the elastic stiffness and the 5% damping as
used in the force-based design or assessment.
The required structure displacement demand is compared to the ultimate displacement capacity.
Acceptable performance is indicated by the ratio sdsc resulting greater than one.
Figure 4. Equivalent stiffness and equivalent damping for the Shibata-Sozen approach [NZSEE 2002]
2.2.3 Comments
The above summaries show that the objective common to both approaches is the determination
of the structure lateral capacity and of the structure yielding and ultimate displacements. For this
purpose, either a simplified hand procedure, i.e. the SLM approach, or a more accurate push-
over analysis may be resorted to. Furthermore, both procedures reduce the structure behaviour
to that of a SDOF oscillator, characterised by a force-displacement relation. Thus, their only
difference lies in the way the earthquake demand on the frame is established, which is however
not a secondary point. The force-based approach uses an acceleration spectrum in conjunction
with the initial elastic period of the structure to estimate the required ductility demand; the
displacement-based approach uses a displacement spectrum and an equivalent period to estimate
the maximum displacement demand.
8/13/2019 Dissertation2003-Lupoi
30/78
16 Chapter 2. Reference Documents
Both criteria are admittedly approximate, and a judgement on their relative merits is not within
the scope of the present study. A comment that is relevant to the present study, however, is that
the magnitude of the difference between the results obtainable with the two criteria can be easily
absorbed in the overall approximation of the SLM method, which varies from case to case and
rests to a considerable extent on the skill of the analyst. The push-over analysis, on the other
hand, which incidentally is a displacement-based procedure, gives a more uniform level of
approximation and, in spite of its own limitations, covers a much more wide range of practical
applications than the SLM method.
2.3 Japanese Guidelines
The first publication of the Standard for Seismic Vulnerability Assessment of Existing
Reinforced Concrete Buildings, edited by a committee organised by the Ministry of
Construction, goes back to 1977. The Standard was revised in December 1990 and its
effectiveness has been tested at the occasion of several earthquakes, for school buildings in
particular.
The standard may be applied to evaluate the seismic performance of an existing reinforced
concrete building whose structural system is made of moment resisting frames, with or without
shear walls. Its application is limited to low-rise buildings, since the standard assumes constant
acceleration response (flat response acceleration spectrum with periods) [Otani 2000].
Three procedures of different accuracy and reliability are available for the assessment. Thefirst-
level procedure is a simple screening procedure aimed at classifying earthquake resistant
buildings based on their storey shear strengths, provided by either columns or/and structural
walls. The buildings classified as questionable by the first-level procedure, must be analysed
by the more sophisticated second-level procedure, in which the deformation capacities of
vertical members are also considered. Finally, the third-levelprocedure is a refined version of
the second level, in which also the weak beam-strong column mechanism is included [FIB
2003]. The attention of the present study is concentrated on the latter procedure.
The Japanese assessment consists in the comparison of the floor shear demand due to the
earthquake action versus the floor shear capacity, for every storey and every frame of the
building:
ic
id VV .
(15)
The floor shear capacity is defined by the following product:
iiic ESTV 00= .
(16)
8/13/2019 Dissertation2003-Lupoi
31/78
Japanese guidelines 17
In Eq. (16), the factor T accounts for the time-dependent deterioration of the members strength
and the factori
S0 accounts for geometry or stiffness irregularity and/or mass concentration at
floor i . Both Tand iS0 are functions of several factors, whose number depends on the level of
the assessment. Their evaluation is based on the results of field-surveys.
The most relevant factor in Eq. (16) is the basic structural performance index iE0 , which is
defined by the following product:
iiU
iqVE =0 ,
(17)
wherei
UV is the shear resistance of all vertical elements at floor i andi
q is the ductility index,
a measure of the deformation capacity of floor i .
The shear resistancei
UV depends on the type of failure mechanism. The latter is established
through simple equilibrium considerations. The lesser between the shear force due to a plastic
hinges (flexural) mechanism and the members shear strengths gives the floor shear resistance
iUV .
The possibility of a weak-beam/strong-column failure mechanism is also investigated. To this
end, a joint index similar to that of New Zealand document may be resorted to:
UcaUcb
UbrUblj
MM
MMS
+=
++ or
UcaUcb
UbrUblj
MM
MMS
+=
+ . (18)
where +UblM is the positive ultimate bending moment of the beam on the left side of the node,
UbrM is the negative ultimate bending moment of the beam on the right side of the node, UcbM
and UcaM are the ultimate bending moment of the column below and above the node,
respectively. If 1jS , a premature beam failure is expected and the shear force on the columns
is limited by the resistance of the beams.
An interesting feature of the Japanese third-level method is about the mixed-type structures, in
which different structural types for the vertical elements are present. It is observed that thepresence of a stiffer and less ductile vertical element may significantly reduce the resistance of
the whole structure. To account for this fact, the expression of the basic performance index in
Eq. (16) is modified as:
( ) 1332210 qVVVE iUiUiUi ++= , (19)
where the shear capacity of the brittle member 1is fully accounted for, while the capacity of the
others are reduced by a coefficient that accounts for the limited deformation capability
8/13/2019 Dissertation2003-Lupoi
32/78
18 Chapter 2. Reference Documents
imposed by member 1. The deformation capacity of the whole structure is thus governed by the
less ductile, i.e. 1qq= .
The s and 1q depend on the type of vertical elements. A classification of them on the basis
of their type of failure is provided in the standard: the eight classes considered are reported in
the table below. All the vertical elements of the structure have to be classified in a maximum of
three groups.
The values for and q are given as functions of the class combination. The corresponding
tables from the Japanese standard are provided below.
Table 1. Classification of vertical elements
Element Definition
Col. m Column whose failure mechanism is in flexureCol. s Column whose failure mechanism is in shear
Col. ss Short column whose failure mechanism is in shear failure and its ratio between
height and depth is smaller than two
Col. mb Column whose ultimate lateral force is caused by flexural yielding of beams
Col. sb Column whose ultimate lateral force is caused by shear failure of beams
Wall m Wall whose failure mechanism is in flexure
Wall s Wall whose failure mechanism is shear
Wall upl Wall whose ultimate lateral capacity is caused by uplift of foundation
Table 2. Value of displacement compatibility factor
Element of the first group
Col. ss Col. s or Wall s
Col. m 0.5 0.7
Col. s 0.7 1.0
Wall m or Wall 0.7 -
Table 3. Ductility index
Mechanism q
Beam Shear 1.5Beam Flexure 3
Column Shear 1
Column Flexure ( )( ) 05.0175.012 + Short Column Shea 0.8
Wall Shear 1
Wall Flexure 1 2
Wall Uplift 3.0
In Table 3, the coefficient is evaluated by:
8/13/2019 Dissertation2003-Lupoi
33/78
8/13/2019 Dissertation2003-Lupoi
34/78
8/13/2019 Dissertation2003-Lupoi
35/78
8/13/2019 Dissertation2003-Lupoi
36/78
22 Chapter 3. Comparative Study
Figure 5. View of Pavia frame
Figure 6. Gravity loads for Pavia frame
The column have square cross-sections with dimension equal to 20cm while the beam have
dimensions of 20x33cm. The geometry and the reinforcement of the frame are shown in the
Figure 7.
8/13/2019 Dissertation2003-Lupoi
37/78
8/13/2019 Dissertation2003-Lupoi
38/78
24 Chapter 3. Comparative Study
Figure 8. Four-storey frame, Icons
Figure 9. Gravity loads for Icons frame
3.2.3 Spear Frame
The last case study refers to a structure built in the framework of the European program SPEAR
[Fardis, 2002] and is a simplification of an actual three-storey building representative of older
construction in Southern Europe. It has been designed for gravity loads alone following the
prescriptions of the Greek design code in use between 1954 and 1995. The storey height is
3.0m.
8/13/2019 Dissertation2003-Lupoi
39/78
Terms of comparison 25
Figure 10. Geometric characteristics of the three test-structures
COLUMNS C1-C5 & C7-C9
STIRRUPS8/25414
1014
STIRRUPS8/25
COLUMN C6
Figure 11. Plant of Ispra frame and details of the columns reinforcement
3.3 Strengths of materials
A reliable estimation of the materials strengths is fundamental in assessment. All codes require
both to use the best estimates of the actual strength instead of nominal values and to account for
the uncertainty of the available information.
The mean values of the tests performed on concrete and steel are assumed as the most probable
strengths; the values used in the applications of the procedures are given in Table 4.
8/13/2019 Dissertation2003-Lupoi
40/78
26 Chapter 3. Comparative Study
Note that in both the Pavia and the Icons frame two concrete qualities were used. In the Pavia frame the
type 2 is referred only to the columns of the first floor, while in Icons frame type 1 is for all the beams
and type 2 for the columns.
Table 4. Materials strength of the test structures (MPa)
Pavia Icons Spear
Concrete 1, fc 14.01 14.28 18.00
Concrete 2, fc 18.53 18.34
Steel, fy 365.7 331.0 255.0
3.4 Terms of comparison (capacity measure and earthquake input model)
The comparison between the selected procedures is made in terms of the PGA value that causes
the collapse of the structure. The PGA has been arbitrarily related to the Type 1 spectrum of
EC8 for a site of soil category C; it is believed that the results of the comparisons would not
change to any significant extent if a different reference spectrum were selected.
The fundamental parameters of the spectrum, whose analytical expression are reported in
Appendix A, are the corner periods BT , CT and DT (see Figure 12), which are respectively
equal to 0.2, 0.6 and 2.0 seconds, and the ratio between the spectral acceleration at the plateau
and the peak ground acceleration, which is equal to 2.875 assuming a site coefficient Sequal to
1.15. In Figure 12 are shown the elastic spectrum for gPGA 1= , %5= and the design
spectrum for 5=q .
Soil C, Type 1
Tb Tc Td0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Period [sec]
agS[g]
Elastic
Design
Figure 12. EC8 elastic and design spectra
The displacement spectrum is related to the elastic spectrum by: ( ) ( )
2
,,
TST ee = .
8/13/2019 Dissertation2003-Lupoi
41/78
Terms of comparison 27
Once the actual spectral acceleration at failure faS , is evaluated, the corresponding PGA at
failure is obtained as:
( ) ( )[ ]scscelfaelelaf TSTSPGA = ,
1,
,, (24)
where ( )elela TS , is the elastic spectral acceleration corresponding to the normalized spectrum,
i.e. for gPGA 1= .
Since the purpose here is not to check whether a structure is able to withstand an earthquake of
given intensity, but to assess its capacity, i.e. the PGA at failure, some minor adjustments are
needed in the practical application of the procedures: they are indicated in the appropriate
places.
3.5 Non-linear numerical analysis
Two-dimensional models of the frames and a three-dimensional model of the SPEAR building
have been set up, employing fibre elements to represent columns and beams. The Manders
constitutive model [Mander et al. 1988] has been adopted for concrete, a bilinear material
relationship for the steel, with the hardening factor for the plastic branch equal to 0.002. The
analyses have been carried out by means of the finite element code OPENSEES [McKenna
1997].
Figure 13. Fiber model
8/13/2019 Dissertation2003-Lupoi
42/78
28 Chapter 3. Comparative Study
3.6 Risk reduction factor
The purpose of the codes for assessment and upgrading of existing buildings is to reduce
earthquake risk to the community. However, since it is generally recognised the impracticability
of bringing all existing buildings up to the standard of the new ones, risk factors are often
adopted, allowing for a reduced protection. This aspect has not been introduced in the present
study, whose focus is on the relative efficiency of different approaches in providing estimates of
the actual risk.
8/13/2019 Dissertation2003-Lupoi
43/78
4 EXAMPLE APPLICATIONS
The application of each of the selected procedures to the Pavia frame is described in detail in
this section. For the other two structures only the results are given, in the following chapter 5.
4.1 FEMA 356: linear static
The effects due to the gravity loads and to the earthquake loads are evaluated separately and
then combined.
The evaluation of the seismic action that causes failure of the frame is carried out in two steps: a
lateral load of unit value, 1V , is applied to the frame and the corresponding demandsi
D1 on all
considered members/mechanisms are evaluated. Then, for each mechanism, the lateral load
multiplier which satisfies the corresponding limit state equation is evaluated from:
iiiG
i DDC 1+= , (for member/mechanism i)(25)
wherei
C is the member capacity andiGD is the demand on the member due to gravity loads.
The collapse of the frame is identified by the minimum of over all mechanisms:
{ }ii
f = min .
The spectral acceleration at failure is simply evaluated from Eq. (1), where the elastic lateral
force is given by:
1VV f = . (26)
All the members of the frame are classified as primary elements, i.e. essential to the resistance
of the frame against the seismic lateral load. The members flexural capacities are evaluated
using the probable material strength (see section 3.3) in accordance to the procedures of ACI
318-99; the maximum tensile deformation of the longitudinal bars is set equal to 5%. No
8/13/2019 Dissertation2003-Lupoi
44/78
30 Chapter 4. Example applications
reduction on member flexural strength, which is required if the shear demand exceeds
grossc Af '
6 , is necessary for the frame under examination.
The shear capacities of beams, columns and joints are calculated using the expressions provided
in the Prestandard, which take into account the contributions of both concrete and lateral
reinforcement. A lower-bound of the material strength is used in the calculation of the ultimate
shear capacity, since this action is classified as force-controlled.
The demands on members, iD s, are calculated by means of a linear elastic finite element
analysis. In setting up the model, the stiffness of beams and columns is reduced by factor equal
to 0.5 with respect to the un-cracked value, as explicitly required by the FEMA.
The seismic action 1V is vertically distributed at the floor levels in accordance with the
coefficients VxC :
11 VCF Vxx = , (27)
=
=n
i
kii
kxxVx hwhwC
1
, (28)
where iw is the portion of the total weight lumped at floor x, xh is the height of floorxfrom the
base. The exponent kis equal to 0.1 for sec5.0T , to 0.2 for sec5.2T and varies linearly
in between. The values of the coefficients VxC are shown in Figure 14: the load distribution
along the frame height is almost triangular. The horizontal actions are applied in both +X
direction and X direction.
0.43
0.38
0.19 1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
7
8
9
10
11
12
13
14
15
16
17
18
1 2 3 4
5 6 7 8
9 10 11 12
Figure 14. Numbering of joints and sections and values of the coefficientsof the vertical distribution of the lateral load
8/13/2019 Dissertation2003-Lupoi
45/78
FEMA 356: linear static 31
The failure mechanisms investigated are: flexure and shear failures at the end sections of all
beams and all columns; shear failure in all joints.
The flexural failures are governed by the following limit state equation:
1MMMm GC = . (29)
The values of the local ductility factor m are tabled in the Prestandard as function of the
Performance Level, of the Component Type, of the amount of transverse reinforcement and of
the levels of axial and shear forces acting on the members. The selected values are about 4 for
columns and about 7 for beams.
The shear failures are governed by the following limit state equation:
JCCC
VVV GC
=
321
1 , (30)
where the coefficients 1C , 2C , 3C , J , defined in section 2.1, are equal to 1.0, 1.0, 0.9 and 2.0,
respectively.
The minimum value of the multiplier is reported in Table 5, for each of the failure
mechanism investigated.
It is assumed that the collapse of the structure occurs for the smallest among those listed in
Table 5; which corresponds to the shear failure of joint 1 (see Figure 14). Thus, the elastic base
shear at failure is equal to 124.4 kN.
Table 5. Values of minimum multiplier for each limit states
Shear MomentMechanism
Beam Column Joint Beam Column
172.7 204.8 124.4 372.2 202.9
The (elastic) spectral acceleration at failure is calculated as:
gWCCCC
VS
m
Ef
fa 508.0321
1
, =
=
=
where the values of the coefficient 1C , 2C and 3C , are the same as above and the coefficient
mC is equal to 1.
The period of the structure from an eigenvalue analysis is equal to sec57.0=elT . Finally, the
failure peak ground acceleration is evaluated by:
( )gS
TSPGA fa
elela
f 177.0875.2
508.01,
,
=== .
8/13/2019 Dissertation2003-Lupoi
46/78
4.2 FEMA 356: Nonlinear static
The seismic capacity of the structure is assessed by carrying out a push-over analysis; the lateral
loads are increased until failure occurs in one of the monitored sections or joints. In this case the
format of the limit state equation is the same for force-controlled actions and deformation-
controlled actions, and it consists on a direct comparison between capacity and demand, this
latter being correctly provided by the analysis, in terms either of forces or of deformations. In
this application, a fibre element formulation is used to model both beams and columns, hence
use of local ductility factors or reduction coefficients, as in Eq. (2) and in Eq. (29), is not
necessary.
The flexural capacity is expressed in terms of plastic rotation, c . These are tabled in the
Prestandard as function of the ratio between the amount of transverse reinforcement and the
cross section area, of the level of the shear forces acting on the member, of the amount of the
transverse reinforcement and, for columns, of the level of the axial force acting on member.
Their values vary from 0.002 to 0.02 for the columns and from 0.025 to 0.05 for the beams; the
average values adopted are around 0.020 and 0.026 for columns and beams, respectively.
The shear capacities are evaluated using the same expressions as for the linear procedure.
The horizontal load is vertically distributed as shown in Figure 15: (1) a uniform distribution
and (2) a triangular distribution. They are applied in the two opposite directions, +X and X. A
total of four push-over analyses are performed.
Figure 15. Uniform and triangular vertical distributions of lateral loads
8/13/2019 Dissertation2003-Lupoi
47/78
FEMA 356: Non linear static 33
The capacity/demand check on all sections and all joints is carried out at each step of the
analysis.
The results of the analysis are summarised in Table 6, where U is the displacement at failure
of the control node, which is by definition the node located at the centre of mass of the top
floor, and UV is the base shear at failure. The element and the type of failure are also indicated.
The worst condition among those investigated are cases number 3 and number 4, to which the
smaller ultimate displacement corresponds; the base shear capacity is about 50 kN.
The push-over curve corresponding to case number 3 is bi-linearised to evaluate the yielding
displacement and the effective stiffness of the frame whose values are:
mY 021.0= , mkNKeff 2908= ;
where, as indicated by the Prestandard, the effective stiffness has been taken equal to the secantstiffness calculated at a base shear force equal to 60% of the effective yield strength of the
structure.
Table 6. Collapse point for the four push-over analyses
Case
number
Lateral
load typeDir. U [m] UV [kN] Mechanism
1 (1) X 0.070 51.6 Flexural failure of column 3
2 (1) -X 0.070 51.4 Flexural failure of column 2
3 (2) X 0.063 51.7 Flexural failure of column 2
4 (2) -X 0.063 51.5 Flexural failure of column 2
The period is evaluated taking into account the elongation effect through the expression:
sec68.0==eff
ineleff
K
KTT ,
where the initial stiffness, mkNKin 4132= , is the elastic lateral stiffness of the building in the
direction under consideration.
The spectral acceleration which causes the collapse of frame is:
ggTCCCC
Seff
ufa 380.0
42
2
3210
, =
=
,
with 0C , 1C , 2C and 3C taken equal to 1.2, 1.0, 1.2 and 1.0 respectively. The corresponding
failure PGA is:
( )g
TS
SPGA
effela
fa
f 150.0543.2
380.0
,
,=== .
8/13/2019 Dissertation2003-Lupoi
48/78
34 Chapter 4. Example applications
It is noted that the displacement ductility capacity of the frame estimated from the push-over
analysis, i.e.:
0.3=
=
y
usc ,
is not explicitly used in the evaluation of faS , and hence of PGAf. This is a feature common to
all displacement-based approaches which adopt the equal displacement rule.
4.3 NZ: Simple lateral Mechanism Analysis
As already pointed out in section 2.2.3, the differences between the force-based and the
displacement-based procedures consist essentially in the way the demand on the structure is
calculated, i.e. either by using an acceleration spectrum or a displacement spectrum. In fact, the
evaluation of the structural capacity, which represents the main task of the assessment, is
common to both procedures. Accordingly, in the application instead of following the sequence
of steps outlined in the NZ document, for each of the two procedures, the evaluation of the
frame capacity has been carried out first, using both the SLM method and the push-over
analysis. The two approaches (i.e. force and displacement based) have then been separately
applied for the assessment of the PGA at failure.
The ultimate flexural capacitiesU
M of members are evaluated by the standard theory [Park and
Paulay, 1975] assuming the maximum tensile strain of the reinforcement bars equal to 0.10, at a
difference with the FEMA value of 0.05. The columns capacity UM is calculated taking into
account the earthquake-induced axial forces, which are estimated through simplified
considerations based on global equilibrium between external and internal actions. Furthermore,
a reduction of UM due to the P effects is applied. The ultimate shear strengths, CUV , of
beams, columns and joints are evaluated by means of the expressions provided in the document.
The degradation of shear strength due to inelastic deformation associated to plastic-hinges is not
accounted for at this stage.
The identification of the most-likely failure mechanism is carried out as follows. First, the type
of failure at the end sections of beams and columns is assessed by comparing the shear demands
on members generated by the plastic hinge (flexural) mechanism, flV , with the shear strengths
CUV . The condition flCU VV > is found to be satisfied at all members sections, therefore
flexural failures are expected.
8/13/2019 Dissertation2003-Lupoi
49/78
NZ: Simple lateral Mechanism Analysis 35
The sway indexes are then calculated according to the Eq. (7). They are given in Table 7 for
both directions of application of the ground motion. Since the values are all larger than 0.85, a
failure mechanism of column-sway type at first floor is predicted.
Table 7. Sway Indexes
Si 1stfloor 2
ndfloor
X dir. 2.03 1.73
-X dir. 2.21 1.95
In this case the probable lateral load capacity can be simply evaluated by means of equilibrium
conditions: the maximum base shear UV is taken equal to the sum of the ultimate flexural
capacities at top and bottom of the columns of the first floor divided by the storey height. The
values of UV are given in Table 8 together with the probable base shear at yielding, calculated
using the yielding flexural capacities of members. The lower values of UV with respect to YV
are due to the P effect.
Table 8. Base Shear at yielding and collapse
Base Shear YV UV
X dir. 45.9 37.9
-X dir. 50.9 43.9
A simplified hand approach is also followed for the determination of the available (global)
displacement ductility of the frame. The (local) ductility capacities in curvature c are
evaluated first by means of sectional analysis. The expressions provided in the NZ document
might have been used alternatively. The values of c of columns at the first floor are given in
Table 9: the expected available curvature ductility is in the order of 10.
On the basis of the estimated c , the reduced shear strengths are evaluated by means of the
degradation law proposed by Priestley et al. included in the document. The values of the
reduced shear strengths crV for the columns at the first floor are also given in Table 9. They are
compared with flV : the condition flcr VV > is satisfied. Thus, the reduction of the (local)
curvature ductility capacities is not necessary.
8/13/2019 Dissertation2003-Lupoi
50/78
36 Chapter 4. Example applications
Table 9. Curvature ductility and shear demand for the four base-columns
Section 7 10 13 16
c 11.4 12.0 9.8 11.6
crV 17.2 18.4 18.1 16.9flV 11.1 12.2 11.9 10.7
The (global) inelastic displacement capacity is evaluated as the sum of the yield and the plastic
components.
The frame yielding displacement is evaluated by the simplified relation Eq. (11) proposed by
Priestley:
mmhh
le
b
byy 164371
330
13400018.05.05.0 =
=
The estimated displacement capacity of the frame is calculated using Eq. (9) and Eq. (10) as:
mmpyu 553916 =+=+= .
The global displacement ductility of the frame is equal to:
48.3=
=
y
usc
The evaluation of the capacity of structure is thus completed; the corresponding bilinear
approximation of the force/displacement curve is shown in Figure 16.
The seismic performance is now determined by either a force or a displacement based approach.
In the force approach, the period is evaluated by an eigenvalue analysis, sec57.0=elT , and one
obtains for the peak ground acceleration at failure:
( )g
M
V
TSPGA Usc
elela
f 204.075.27
9.3748.3
875.2
11
,
=== .
8/13/2019 Dissertation2003-Lupoi
51/78
NZ: Simple lateral Mechanism Analysis 37
0 10 20 30 40 50 600
10
20
30
40
50
60
displacement [mm]
Force,
V[kN]
Edx
Esx
Figure 16. Resultant SDOF from mechanism analysis
0.1690.204
0.586
0.00
0.25
0.50
0.75
0.0 0.5 1.0 1.5 2.0 2.5
Period [sec]
Sa
[g]
elastic
design q=3.48
Figure 17. Force approach of SLM for the Pavia Frame
In the displacement approach, the characteristics of the substitute structure are:
sec26.12 ==
eff
eff
k
MT
( ) %20, == mechfailuref sceq .
The PGA at failure is:
( )g
TSPGA effu
eqeffela
f 161.0,
1 2
,
==
In the above expression, elaS , is the normalized elastic spectrum evaluated for the damping
coefficient eq using the reduction factor 63.05
10=
+=
eq .
8/13/2019 Dissertation2003-Lupoi
52/78
4.4 NZ: Nonlinear Static
The NSP procedure is intended as a refinement of the SLM analysis: in particular, the lateral
force capacity and the available displacement ductility of the structure are evaluated through a
push-over analysis instead of using the hand approach illustrated in the previous section.
The calculation of members capacities both in flexure and in shear is carried out as for the SLM
case. Therefore, no further comments are necessary here, except that the flexural failure is
evaluated in term of curvature deformations instead of bending moments. The ultimate
curvature capacities of the sections are derived from sectional analysis as before: for the
columns of the Pavia frame, the corresponding values vary between 0.2 and 0.3 mm-1.
The finite element model of the frame has been described in section 3.5. The two horizontal
loads distributions specified by FEMA have been applied also for this case, since no explicit
indication on this subject is provided in the NZ-document.
The four curves obtained by combining the two directions of application and the two vertical
distributions of horizontal loads are shown in Figure 18. The probable lateral load capacity is in
the order of 50 kN.
PAVIA Frame
0
10
20
30
40
50
60
0 0,02 0,04 0,06 0,08 0,1 0,12
Displacement [m]
BaseShear[kN]
Triangular distribution
Uniform Distribution
Figure 18. Push-over curves for two types of distributions
To establish the available displacement ductility of the frame, in the NZ document it is
suggested to push the frame further, deforming as a mechanism, until the available ultimate
curvature is reached at the critical plastic hinge. The corresponding ultimate displacement is
read directly from curve.
8/13/2019 Dissertation2003-Lupoi
53/78
NZ: Nonlinear static 39
In the case under consideration, the flexural capacity is exhausted at the bottom of the external
column 4, for the +X direction of application, and at the bottom of the opposite external column
2, for the -X direction of application.
The reductions of shear strengths due to local inelastic deformations are evaluated next, in the
same way as described in the previous section. It is noted, however, that the NZ document
requires to perform this check at failure, i.e. to use the values of the local inelastic deformation
and of the flexural shear demand flV corresponding to the frame configuration at failure. If
the check is negative, the ultimate deformation is reduced to that pertaining to shear failure.
In two out of the four cases investigated, shear failures are observed well before the flexural
ultimate capacity is reached. The corresponding points in the force/displacement curve are
shown in Figure 19: it is noted, however, that the two shear failures occur when the lateral force
has practically reached the ultimate capacity, implying that this result is quite sensitive to
accuracy of the shear strength model used.
PAVIA Frame
0
10
20
30
40
50
60
0 0,02 0,04 0,06 0,08 0,1 0,12
Displacement [m]
BaseShear[kN]
Triangular distribution
NZ shear failures
NZ flexural failures
Figure 19. Shear and flexural failures for the triangular distribution
The significant quantities derived from the push-over analyses are summarized in
Table 10: while the lateral load capacity is almost equivalent for all the cases, the reduction in
the available ductility due to shear failure in the two cases discussed above is considerable.
The values of the yielding displacement given in the table have been derived after the
bilinearisation of the push-over curves.
8/13/2019 Dissertation2003-Lupoi
54/78
40 Chapter 4. Example applications
Table 10. Collapse point for the four analyses
Shape Dir. UV Y U sc mechanism
(1) X 50.0 0.023 0.034 1.52 Shear failure of joint 4
(1) -X 50.6 0.023 0.036 1.55 Shear failure of joint 1
(2) X 51.1 0.021 0.106 5.03 Flexural failure of section 10
(2) -X 50.9 0.021 0.101 4.91 Flexural failure of section 10
As for the SLM analysis, the failure PGA is calculated following both a force approach and a
displacement approach.
In the force approach, the peak ground acceleration at failure is:
( )g
M
V
TSPGA Usc
elela
f 091.075.27
0.5052.1
875.2
11
,
=== .
0.261
0.091
0.171
0.00
0.10
0.20
0.30
0.0 0.5 1.0 1.5 2.0 2.5
Period [sec]
Sa
[g]
elastic
design q=1.52
Figure 20. Design and elastic spectra for PGA at failure
In the displacement approach, the characteristics of the substitute structure are:
sec866.02 ==eff
effk
MT
where mkNkeff /1460034.0
0.50== and %8.502.0
112.0 =+
=sc
eq
.
The PGA at failure is:
( )g
TSPGA effu
eqeffela
f 096.0,
1 2
,
==
In the above expression, elaS , is the normalized elastic spectrum evaluated for the damping
coefficient eq using the reduction factor 96.0= .
8/13/2019 Dissertation2003-Lupoi
55/78
4.5 BDPA: 3rd
level
The first step in the application of the Japanese procedure consists in the classification of the
vertical elements into the eight classes available. In the Pavia frame, all the vertical elements are
columns having the same section and reinforcement layout. To establish the type of failure of
each column, their flexure and shear strengths are calculated, at all floors, by means of the
expressions provided in the guidelines. From the equilibrium of the ultimate bending moment,
the shear forces on members associated with a plastic-hinges (flexural) mechanism are derived
and compared with the ultimate shear. It results that all the columns at all floors fail in flexure,
i.e. flCU VV > .
The possibility of beams failure is investigated by means of the joint-indexes, which are
calculated with the expressions in Eq. (18). Values greater than 1 are found at all nodes. Thus, a
strong-beam/weak-column type of failure mechanism is expected for the Pavia frame. It is
concluded that all the vertical elements belong to class m: columns whose failure mechanism is
flexural yielding.
In this case, the expression of the basic structural performance index defined in Eq. (17) reduces
to:
( )==nc
j
iUjj
i VqE1
0 min , (31)
where iUjV is the shear capacity of column jat floor i , nc is the number of columns at that floor
and jq is the ductility index of column j . The latter depends on the coefficient defined in
Eq. (20), which for the case under examination is the same for all columns at all floors.
The floor shear resistance is evaluated by the expression:
( )=
=nc
j
iUjj
iic VqTSV
1
0 min , (32)
For the Pavia frame the two indexesi
S0 and Tare both equal to 1 because the frame has a
regular layout and it is a new construction, i.e. the materials are not deteriorated. The values
ofi
cV are given in Table 11.
According to Eq. (23), the failure spectral acceleration at floor iis calculated as:
i
ii
akm
ES= 0 , (33)
8/13/2019 Dissertation2003-Lupoi
56/78
42 Chapter 4. Example applications
The values of iaS are given in Table 11. The collapse of the frame occurs for the smallestiaS ,
which is equal to 0.477g.
In Eq. (33) the coefficients ik are equal to 0.86, 1.07 and 1.28 respectively for the first, the
second and the third floor.
Table 11. Results of the 3rdlevel procedure for the three different floors
Storey Mechanism q Vu[kN] E0[kN] Sa [g]
1 Col. Flexural 3.2 50.5 151.4 0.649
2 Col. Flexural 3.2 43.8 140.2 0.775
3 Col. Flexural 3.2 37.0 118.6 1.402
The PGA is given by the expression in Eq. (24). The elastic period, evaluated from the
eigenvalue analysis is equal to: sec57.0=elT . The factor ( )ela TS1 is equal to 2.875. Finally,
the peak ground acceleration at failure is equal to 0.226g.
8/13/2019 Dissertation2003-Lupoi
57/78
8/13/2019 Dissertation2003-Lupoi
58/78
44 Chapter 5. Comparisons
Icons Frame - Basic-Procedures
0.2290.209
0.253
0.298
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
NZSEE MA
Force
NZSEE MA
Disp
FEMA LSP BDPA
PGA[g]
Pushover Procedures
0.113
0.191
0.140
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
NZSEE
NPA Force
NZSEE
NPA Disp
FEMA
NSP
Figure 22. Failure PGA for the Icons frame
The variation of failure PGA s which results from the two push-over procedures is, on the
contrary, unexpected. Since the model of the frames and the external action applied are the
same, the difference in the results has to be related either to the capacities of members or to the
way the push-over curve is related to the PGA at failure. The points on the push-over curves
which correspond to the exhaustion of capacity of the indicated member are shown in Figure 23
and Figure 24.
PAVIA Frame
shear joint 4
shear joint 1
flexure col 4
flexure col 4
flex col 3
flex col 2flex col 2
flex col 2
0
10
20
30
40
50
60
0 0.02 0.04 0.06 0.08 0.1 0.12
Displacement [m]
BaseShear[kN]
Triangular distribution
Uniform Distribution
NZ failures
FEMA failures
Figure 23. Collapse points of push-over procedures for Pavia frame
8/13/2019 Dissertation2003-Lupoi
59/78
45
ICONS Frame
flexure col 2
flexure col 5
shear col 2
shear joint 8