FRAGILITY CURVES FOR BUILDING STOCKS: DERIVATION FROM...
Transcript of FRAGILITY CURVES FOR BUILDING STOCKS: DERIVATION FROM...
FRAGILITY CURVES FOR BUILDING STOCKS: DERIVATION FROM MACROSEISMIC AND MECHANICAL-BASED MODELS AND
CALIBRATION THROUGH EMPIRICAL DATA
Serena Cattari
DICCA–Department of Civil,Chemical andEnvironmental Engineering
29th October 2019,LIMA
OUTLINE OF THE PRESENTATION
What do theydepend on?
How are theyobtained?
What do theyrepresent?
How can they beused?
Fromvulnerability torisk assessment
Practical issues & Examples of application
Overview & focus on macroseismicand mechanical-based approaches
Involved uncertainties & influenceon results
Relationship with risk analyses &useful GLOSSARY ….
OUTLINE OF THE PRESENTATION
What do theydepend on?
How are theyobtained?
What do theyrepresent?
How can they beused?
Fromvulnerability torisk assessment
Practical issues & Examples of application
Overview & focus on macroseismicand mechanical-based approaches
Relationship with risk analyses &useful GLOSSARY ….
Involved uncertainties & influenceon results
RISK ANALYSIS - KEYWORDS
HAZARDASSESSMENT VULNERABILITY
& DAMAGE ESTIMATIONEXPOSURE &
CLASSIFICATION
LOSS ESTIMATION
Loss
The risk analysis at territorial scale is intrisically probabilistic (as the PEER-PBA at scale of the single building): it is the result of convolution of various sources of uncertainties and dispersions
RISK ANALYSIS - BASICS
It represents the EXPECTED RATE in a GIVEN TIME (e.g. 1 year, 50 year,…) of possible LOSSES ( economical, associated to the loss of buildings usability, casualties, …) due to the DAMAGE
OCCURRED on the buildings stock or people (EXPOSURE) in a GIVEN AREA (e.g. the municipality, the region, the whole country) as a consequence of possible seismic events (HAZARD)
FRAGILITY
A chain of conditional probabilities….
Notation oftheintegral according toIervolino,2016
RISK ANALYSIS – HAZARD
Hazard function (PSHA - Probabilistic Seismic Hazard Assessment) stands for the probability that a selected scalar measure of seismic intensity at a site exceeds a given value in a given time interval
PGA o Sa 16- 50-849 TR
PHSA provided by the Italian Structural Code
0.000#
0.001#
0.010#
0.100#0.000# 0.500# 1.000# 1.500#
Sa(T1)m#
Sa#(T1)#16%#
Sa#(T1)#84%#
ag#
ag16%#
ag84%#Foragiven site
RISK ANALYSIS –TYPES OF RISK ANALYSISaccording to the notation introduced by FEMA P-58 for performance assessment of buildings
Evaluates the loss over a specifiedperiod of time
(e.g., 1 year, 30 years, or 50 years) considering all earthquakes that
could occur in that time period, and the probability of occurrence
associated with each earthquake.
TIME-BASED INTENSITY-BASED SCENARIO-BASED
The timeperiod depends on the interestsand needs of the decision-maker.Assessments based on a single year are useful for cost-benefit
evaluations, instead over longerperiods of time are useful for other decision-making aims.
UNCONDITIONED
Evaluates the lossassuming a scenario
consisting of a specific magnitude
earthquake occurringat a specific location
relative to the area under examination
CONDITIONED
Useful to support the design of seismic
emergency plan atmunicipality scale
Evaluates the loss assumingthat of a specified
earthquake shakingintensity. In general the
design earthquake shakingconsistent with a Building
Code responseSpectrum is considered.
CONDITIONED
Useful for comparative analysesby considering RARE (e.g. 475 years) and FREQUENT (e.g. 50
uears ) events
IncaseofCONDITIONEDASSESSMENTITISPOSSIBLEALSOADOPTTHEMACROSEISMICINTENSITYas INTENSITYMEASURE
RISK ANALYSIS –SCALES OF RISK ANALYSIS
The whole country
The region
The municipality
RISK ANALYSIS – DAMAGE METRIC
GRADE 1:Negligible to slight damage
GRADE 2:Moderate damage
GRADE 3:Substantial to heavy damage
GRADE 4:Very heavy damage
GRADE 5:Destruction
DAMAGELEVELACCORDINGTOEMS98(Grunthal 1998)
• Usually, the damage is described in DISCRETE terms rather than as a CONTINOUS variable (à thus the integral becomes a sum)
• It is associated to DAMAGE LEVELS (that usually implicitly refer also to PERFORMANCE LEVELS) correlated to Engineering Demand Parameters representative of the Structural Response
• The most common assumption is referring to the 5 grades introduced by the EMS98 scale (the first four are similar to the LSs of Codes)
Risk analyses at large scale can be referred:
q to stock of buildings characterized by homogeneous seismic behaviour – typicalof the assessment on residential buildings spread on the territory
q to buildings portfolio characterized by a group of structures whose assessmentis provided in aggregate terms – typical of the assessment on strategic functionsspread on the territory (e.g. schools, strategic buildings, …)
Thus important steps of a risk analysis are :
q Taxonomy: aimed to define the attributes that influence the vulnerabilityq Classification: group of buildings with the same attributes (similar behavior)
Both can be defined IN GENERAL but then the attributes to be actually consideredin a specific risk analysis depend on the availability of data.
RISK ANALYSIS –EXPOSURE
LIST OF 13 ATTRIBUTES1. Direction2. Material of the lateral load-resisting system3. Lateral load-resisting system 4. Height5. Date of construction or retrofit6. Occupancy7. Building position within a block8. Shape of the building plan9. Structural irregularity10. Exterior walls11. Roof12. Floor13. Foundation system
Reference:GEM Building Taxonomy Version 2.0GEM Technical Report 2013-02Version: 1.0.0Date: November 2013
FRAGILITY CURVES: TAXONOMY & CLASSIFICATIONEXAMPLES
TAXONOMYLIST OF 13 ATTRIBUTES
1. Direction2. Material of the lateral load-resisting system3. Lateral load-resisting system (LLRS) 4. Height5. Date of construction or retrofit6. Occupancy7. Building position within a block8. Shape of the building plan9. Structural irregularity10. Exterior walls11. Roof12. Floor13. Foundation system
MASONRY BUILDINGSYNER-G project (http://www.vce.at/SYNER-G/)
FRAGILITY CURVES: TAXONOMY & CLASSIFICATION
CATEGORY CLASSIFICATION
FRM
Bearing Walls (BW) Out of plane (OP); In plane (IP) [Equivalent Frame (EF), Weak Spandrels Strong Piers (WSSP), Strong Spandrels Weak Piers (SSWP)]
FRM
M
Unreinforced Masonry (URM)
Reinforced Masonry (RM) Confined Masonry (CM) Timber-framed Masonry
(TM)
Blocks: Adobe (A); Fired brick (FB); Soft Stone (SS); Hard Stone (HS) [Regular Cut (RC), Uncut (UC), Rubble (RU)]; Hollow clay tile (HC) [High % of voids (H%), Low % of voids (L%), Concrete Masonry Unit (CMU), Autoclaved Aerated Concrete (AAC)]
Mortar: Lime mortar (LM); Cement mortar (CM); Mud mortar (MM); Hydraulic mortar (HM)
Strengthening: Strengthened masonry (Sm) Timber: Confined and braced masonry
panels (TC); Horizontal timber tie (TT) Concrete and reinforcement: [Average
Strength (20-50 MPa)(ASC), Low Strength (<20 MPa)(LSC)]; [Vertical Reinforcement Bars (RBV), Vertical and Horizontal Reinforcement Bars (RBVH)]
P Regular (R) Irregular (IR)
[Isolated (I), Aggregate (A)]
E Regular geometry (R) Irregular geometry (IR)
CO
Regular openings (RO) Irregular openings (IRO)
[High % voids (H%), Low % voids (L%)]
DM
Details: High quality details (HQD), Low quality details (LQD)
Maintenance: Good Maintenance (HM), Low Maintenance (LM)
Tie rods: Without tie rods (WoT); With tie rods (WT)]
Ring beams: Without ring beams (WoRB); With ring beams (WRB)
FS
Rigid (R) Flexible (F)
Reinforced concrete (RC); Steel (S); Timber (T); Vault (V)
RS
Peaked (P) Flat (F) Gable End Walls (G)
Material: Timber (Ti); Corrugated Metal Sheet (CMS); Reinforced Concrete (RC); Thatch (Th)
Thrusting roof (Tr); Unthrusting roof (UTr)
HL
Low-rise (1-2) (L) Mid-rise (3-5) (M) High-rise (6-7) (H) Tall (8+) (Ta)
Number of stories (indicate the number)
C Pre-Code (PC) Pre-code Aseismic Construction: Low Level
TAXONOMY CLASSIFICATIONby a string of tags
Each vulnerability class, which can be syntheticallynamed by a number or a short acronym, is clearlyidentified by a precise taxonomy:URM2-M: BW-IP\URM-HS-UC-LM\R\R\x\LQD-WT\F-T\P-T\M\PCURM3-H: BW-IP\URM-FB-LM\R\R\x\LQD-WT\R-S\P-RC\H\PC
m*
h*
m*
h*
XY
1"
2"
3"
4"
5"
DATA PROCESSING TO DEFINE THE CLASSES DEFINITION OF PARAMETERS EVALUATION OF CAPACITY CURVES For each class and both directions X and Y
For the random variables
84%
0"
0,01"
0,02"
0,03"
0,04"
0,05"
0,06"
0,07"
40" 45" 50" 55" 60" 65" 70" 75" 80" 85" 90"
16%
50%
X direction Y direction
Mechanical parameters and loadsτk,x shear strengthGx,i shear modulusγi specific weight
ΔS,LS4 ΔF,LS4
drift limit values for the shear and flexural response
κx,i spandrel contribution on the masses
qi - ζx,i load and orientation of floors
Sd
Sa
Du,x
Au,x
4π2/Ty,x2
Sd
Sa
Du,y
Au,y
4π2/Ty,y2
Geometrical features
hi inter-storey heightαx,i ratio of resistant wall
over the gross area A
Modal shape
φx,i i-th component of theassumed modal shape
Corrective factors
Κ1,Κ2,Κ3 , Κ4
for the evaluationof the yieldingacceleration
Κ5 , Κ6 for the evaluationof the period
Failure mode
εx weight assigned to thein-plane SSWPmechanism
FRAGILITY CURVES: TAXONOMY & CLASSIFICATION
MASONRY BUILDINGSYNER-G project (http://www.vce.at/SYNER-G/)
CATEGORY CLASSIFICATION
FRM
Bearing Walls (BW) Out of plane (OP); In plane (IP) [Equivalent Frame (EF), Weak Spandrels Strong Piers (WSSP), Strong Spandrels Weak Piers (SSWP)]
FRM
M
Unreinforced Masonry (URM)
Reinforced Masonry (RM) Confined Masonry (CM) Timber-framed Masonry
(TM)
Blocks: Adobe (A); Fired brick (FB); Soft Stone (SS); Hard Stone (HS) [Regular Cut (RC), Uncut (UC), Rubble (RU)]; Hollow clay tile (HC) [High % of voids (H%), Low % of voids (L%), Concrete Masonry Unit (CMU), Autoclaved Aerated Concrete (AAC)]
Mortar: Lime mortar (LM); Cement mortar (CM); Mud mortar (MM); Hydraulic mortar (HM)
Strengthening: Strengthened masonry (Sm) Timber: Confined and braced masonry
panels (TC); Horizontal timber tie (TT) Concrete and reinforcement: [Average
Strength (20-50 MPa)(ASC), Low Strength (<20 MPa)(LSC)]; [Vertical Reinforcement Bars (RBV), Vertical and Horizontal Reinforcement Bars (RBVH)]
P Regular (R) Irregular (IR)
[Isolated (I), Aggregate (A)]
E Regular geometry (R) Irregular geometry (IR)
CO
Regular openings (RO) Irregular openings (IRO)
[High % voids (H%), Low % voids (L%)]
DM
Details: High quality details (HQD), Low quality details (LQD)
Maintenance: Good Maintenance (HM), Low Maintenance (LM)
Tie rods: Without tie rods (WoT); With tie rods (WT)]
Ring beams: Without ring beams (WoRB); With ring beams (WRB)
FS
Rigid (R) Flexible (F)
Reinforced concrete (RC); Steel (S); Timber (T); Vault (V)
RS
Peaked (P) Flat (F) Gable End Walls (G)
Material: Timber (Ti); Corrugated Metal Sheet (CMS); Reinforced Concrete (RC); Thatch (Th)
Thrusting roof (Tr); Unthrusting roof (UTr)
HL
Low-rise (1-2) (L) Mid-rise (3-5) (M) High-rise (6-7) (H) Tall (8+) (Ta)
Number of stories (indicate the number)
C Pre-Code (PC) Pre-code Aseismic Construction: Low Level
CATEGORY Taxonomy Data at national scale
Data at regional scale
Direction X - Material of the Lateral Load Resisting System − Material type − Material technology − Material properties
X X X
X - -
O
Lateral Load Resisting System − Type of lateral load-resisting system − System ductility
X X
- -
O
Height − Number of stories above the ground − Number of stories below the grown − Height of the ground floor
X X X
X X -
O Date of Construction X X Code Level - Occupancy X X Building Position within a Block X X Shape of the Building Plan X -
Data at national scale (e.g. ITALIAN CENSUS ISTAT) , Material of the lateral load-resisting system: 1) Masonry, 2) R.C. 3) R.C. with “pilotis”, 4) othersData at regional scale, Type of masonry and % of distribution:1a) irregular (25%), 1b) uncut (55%), 1c) soli brick (20%)
FRAGILITY CURVES: TAXONOMY & CLASSIFICATIONAVAILABILITY OF DATA
CATEGORY Taxonomy Data at national scale
Data at regional scale
Structural Irregularity − Plan − Elevation − Type of irregularity (Cladding and Openings)
X X X
-
(X) X
O O
Maintenance - X Exterior walls X - Roof − Shape − Covering Material − Roof system material − Roof connections (Thrusting)
X X X X
- - - -
O O O O
Floor − System material − System type − Connections
X X X
- - -
O O O
Foundation system X -
AVAILABILITY OF DATA
Data at national scale (e.g. ITALIAN CENSUS ISTAT) , NO informationData at regional scale, Type of masonry and % of distribution:1) Timber floor (25%), 2) Vault (35%), 3) Brick – Iron (20%), 4) R.C. (20%)
FRAGILITY CURVES: TAXONOMY & CLASSIFICATION
The case study of GAIOLEIRO buildings in Lisbon between XIX and XX centuries
Rio Tejo
Betão armadoMisto alvenaria – betãoAlvenaria não armadaEdifícios singularesAnexos
Avenidas Novas
FRAGILITY CURVES: TAXONOMY & CLASSIFICATION
REF:Simoes etal.(2019)Fragility functions fortall URMbuildings around early 20° century inLisbon.Part1&2.InternationalJournalofArchitectural Heritage,online.
Tipo I Tipo II Tipo III Tipo IV
MainFacade
BackFacade Plan view Case of study
REF:Simoes etal.(2019)Fragility functions fortall URMbuildings around early 20° century inLisbon.Part1&2.InternationalJournalofArchitectural Heritage,online.
FRAGILITY CURVES: TAXONOMY & CLASSIFICATIONThe case study of GAIOLEIRO buildings in Lisbon between XIX and XX centuries
Epistemic Uncertainties: geometry, structural details and materials
(…)
• Configuração R/C: habitação ou comércio• Solução paredes de empena: meeiras ou independentes• Paredes de empena: tijolo maciço ou tijolo furado• Paredes interiores: tijolo maciço ou tijolo furado• Paredes divisórias: tijolo furado ou tabique de madeira
LOGIC TREE WITH 32 BRANCHES
Tipo I Tipo II Tipo III Tipo IV
REF:Simoes etal.(2019)Fragility functions fortall URMbuildings around early 20° century inLisbon.Part1&2.InternationalJournalofArchitectural Heritage,online.
FRAGILITY CURVES: TAXONOMY & CLASSIFICATIONThe case study of GAIOLEIRO buildings in Lisbon between XIX and XX centuries
➟ from 32 to 8 branches
REF:Simoes etal.(2019)Fragility functions fortall URMbuildings around early 20° century inLisbon.Part1&2.InternationalJournalofArchitectural Heritage,online.
FRAGILITY CURVES: TAXONOMY & CLASSIFICATION
PRELIMINARY NUMERICAL ANALYSES TO OUTLINE THE SIMILARITIES IN THE SEISMIC BEHAVIOUR:
H-S-SH-T-T0.070
H-S-S-S-H0.039
H-I-S-S-H0.078
H-I-SH-T-T0.143
S-S-S-S-H0.078
S-S-SH-T-T0.143
S-I-S-S-H0.159
S-I-SH-T-T0.290
The case study of GAIOLEIRO buildings in Lisbon between XIX and XX centuries
𝑓"#|%# 𝑑𝑚 = 𝑓"# 𝑖𝑚 = 𝑃 𝑑𝑚 ≥ 𝐷𝑀|𝑖𝑚 = 𝑃 𝐼𝑀"# < 𝑖𝑚 = Φ𝑙𝑜𝑔 𝑖𝑚
𝐼𝑀"#𝛽"#
RISK ANALYSIS – FRAGILITY
! "#$|&$ '( "&$ *( '(*()-.
If the damage is described by a CONTINOUS variable
If the damage is described by a DISCRETE variable
OUTLINE OF THE PRESENTATION
What do theydepend on?
How are theyobtained?
What do theyrepresent?
How can they beused?
Fromvulnerability torisk assessment
Practical issues & Examples of application
Overview & focus on macroseismicand mechanical-based approaches
Relationship with risk analyses &useful GLOSSARY ….
Involved uncertainties & influenceon results
FRAGILITY FUNCTIONSThe fragility function of a building class gives the probability that a Damage Levels (DM) is reached given a value im of theIntensity Measure (IM) :
where: IMDM is the median value of the lognormal distribution of the intensity measure imDM for which the DM is attained and βDM
is the dispersion.
IMLS
βLS
βDM
IMDM
𝛽"# = 12𝑙𝑜𝑔 𝐼𝑀89 − 𝑙𝑜𝑔 𝐼𝑀;<
the more the building class is homogenous and the less is b
For IM different possible choicesthe more IM is representative
and the less is b
𝑓"#|%# 𝑑𝑚 = 𝑓"# 𝑖𝑚 = 𝑃 𝑑𝑚 ≥ 𝐷𝑀|𝑖𝑚 = 𝑃 𝐼𝑀"# < 𝑖𝑚 = Φ𝑙𝑜𝑔 𝑖𝑚
𝐼𝑀"#𝛽"#
λ16" λ"
λ84"
λ(IMLS)
IMLS
IMH,84[λ(IMLS)]
IMH,16[λ(IMLS)]
λ"
IM
𝛽"# = 𝛽=>?>@AB + 𝛽@DEB + 𝛽A>F>GDHDIDHB + 𝛽E>J>EKLMB�
PGAwith10%ofexceedance probability
16%quantile 50%quantile 84%quantile
Source:
FRAGILITY FUNCTIONS – involved uncertainties
IMH,16[λ(IMLS)]
Sa1(Sd)
IMH,84[λ(IMLS)]
IMLS
SP
EC
TRA
L A
CC
ELE
RAT
ION
SPECTRAL DISPLACEMENT
𝛽"# = 𝛽=>?>@AB + 𝛽@DEB + 𝛽A>F>GDHDIDHB + 𝛽E>J>EKLMB�
!
0
200
400
600
800
1000
1200
1400
1600
1800
0 0,5 1 1,5 2 2,5 3
Sa (c
m/s
2)
T (s)
AQA_NS
AQA_WE
AQG_NS
AQG_WE
AQK_NS
AQK_WE
AQV_NS
AQV_WE
AQU_NS
AQU_WE
0
2
4
6
8
10
12
14
16
18
20
0 0,05 0,1 0,15 0,2 0,25 0,3
Sa[m/s2]
Sd[m]
AQA-WE
AQA-NS
AQK-WE
AQK-NS
AQV-WE
AQV-NS
AQG-WE
AQG-NS
AQU-WE
AQU-NS
0
0,2
0,4
0,6
0,8
1
1,2
1,4
1,6
0 0,05 0,1 0,15 0,2 0,25
Sa[g]
Sd[m]
MEANVALUE(AQA,AQK,AQV,AQG,AQU)
Sa- Tacceleration spectrum format
ADRSacceleration spectrum formatL’Aquila2009event – recordings fromthe10available stations onthearea
FRAGILITY FUNCTIONS – involved uncertainties
𝛽"# = 𝛽=>?>@AB + 𝛽@DEB + 𝛽A>F>GDHDIDHB + 𝛽E>J>EKLMB�
Numerical curve
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
0.00 0.01 0.02 0.03 0.04
Sa [
m/s
2 ]
Sd [m]
Sd [m]
Sa [m
/s2 ]
Capacity curve
Variability of parameters (mechanical, geometric, …) representative of the behavior of the SINGLE buildings
or the WHOLE CLASS OF BUILDINGS STOCK in accordance with the purposes of a regional loss
assessment
fm - MPa0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
P
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
fm [Mpa]E - MPa
200 300 400 500 600 700 800 900 1000
P
0
0.5
1
1.5
2
2.5
3
3.5
E [Mpa]
Exposure analysis1Single structure at a givensite
Classes of buildings with ahomogeneous seismicbehavior
m*
h*
m*
h*
XY
1"
2"
3"
4"
5"
DATA PROCESSING TO DEFINE THE CLASSES DEFINITION OF PARAMETERS EVALUATION OF CAPACITY CURVES For each class and both directions X and Y
For the random variables
84%
0"
0,01"
0,02"
0,03"
0,04"
0,05"
0,06"
0,07"
40" 45" 50" 55" 60" 65" 70" 75" 80" 85" 90"
16%
50%
X direction Y direction
Mechanical parameters and loadsτk,x shear strengthGx,i shear modulusγi specific weight
ΔS,LS4 ΔF,LS4
drift limit values for the shear and flexural response
κx,i spandrel contribution on the masses
qi - ζx,i load and orientation of floors
Sd
Sa
Du,x
Au,x
4π2/Ty,x2
Sd
Sa
Du,y
Au,y
4π2/Ty,y2
Geometrical features
hi inter-storey heightαx,i ratio of resistant wall
over the gross area A
Modal shape
φx,i i-th component of theassumed modal shape
Corrective factors
Κ1,Κ2,Κ3 , Κ4
for the evaluationof the yieldingacceleration
Κ5 , Κ6 for the evaluationof the period
Failure mode
εx weight assigned to thein-plane SSWPmechanism
REGI
ONAL
SCA
LESI
NGLE
BUI
LDIN
G
Taxonomy | Inventory
FRAGILITY FUNCTIONS – involved uncertainties
𝛽"# = 𝛽=>?>@AB + 𝛽@DEB + 𝛽A>F>GDHDIDHB + 𝛽E>J>EKLMB�
Sd
Simplified Bilinear Curve
Say
SduSdy
?
?
? ?
0.00
0.50
1.00
1.50
2.00
2.50
3.00
0.00 0.01 0.02 0.03 0.04
Sa [
m/s
2 ]
Sd [m] Sd [m]
Sa [m
/s2 ]
Sd / Sdy
0.00
0.20
0.40
0.60
0.80
1.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Sa/S
ay
Sd/Sdy
Sa /
Say
Normalized Bilinear curve
negligible DL3DL1 DL2 DL4
FRAGILITY FUNCTIONS – involved uncertainties
?
im
FRAGILITY FUNCTIONS – DAMAGE PROBABILITY MATRIX (DPM)
DAMAGE PROBABILITY MATRIX represents a typical results (à the fragilityterm) of a CONDITIONED ASSESSMENT
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
βD3 = 0.6
IMD1 = 0.1gIMD5 = 0.9g
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g 0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
FRAGILITY FUNCTIONS– influence of b
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
βD3 = 0.9
IMD1 = 0.1gIMD5 = 0.9g
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g 0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
βD3 = 0.6
IMD1 = 0.1gIMD5 = 0.9g
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g 0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
βD3 = 0.3
IMD1 = 0.1gIMD5 = 0.9g
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
FRAGILITY FUNCTIONS– influence of b
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
IMD1 = 0.15gIMD4 = 0.43g
βD3 = 0.6
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g 0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
FRAGILITY FUNCTIONS– influence of the ductility
00,10,20,30,40,50,60,70,80,9
1
0 0,2 0,4 0,6 0,8
Prob
abili
ttà
PGA [g]
IMD3 = 0.3g
IMD1 = 0.07gIMD4 = 0.64g
βD3 = 0.6
0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.2g0,0
0,1
0,2
0,3
0,4
0,5
0,6
D0 D1 D2 D3 D4 D5
Prob
abilità
im = 0.4g
OUTLINE OF THE PRESENTATION
What do theydepend on?
How are theyobtained?
What do theyrepresent?
How can they beused?
Fromvulnerability torisk assessment
Practical issues & Examples of application
Overview & focus on macroseismicand mechanical-based approaches
Relationship with risk analyses &useful GLOSSARY ….
Involved uncertainties & influenceon results
q Empirical / Observationalq Expertise-based / Heuristicq Mechanical-basedq Hybrid methods
FRAGILITY CURVES: how are they obtained?
All of them pose various critical issues on:
• the incompleteness/reliability of empirical data (Empirical/Observational/Euristic)• the definition of a robust METRIC of DAMAGE• the representativeness of archetype buildings (Mechanical-based)• the need of calibration & validation• the difficulties on defining proper relationships to relate damage to consequence functions• ……
?REF for the classification of methods: Rossetto T., D’Ayala D., Ioannou I., Meslem A. (2014) Evaluation of existing fragility curves , Chapter 3 In SYNER-G: Typology Definition and Fragility Functions for Physical Elements at Seismic Risk: Elements at Seismic Risk, Geotechnical, Geological and Earthquake Engineering 27 pp. 420
FRAGILITY CURVES: how are they obtained?
q Empirical / Observational
ü Derived from observed damage after the occurrence of an earthquakeü Empirical data are usually referred to macroseismic intensity, which is not
an instrumental measure but is based on a subjective evaluation but inrecent experiences are also derived directly in terms of PGA thanks tothe use of shake-maps
ü Data are valuable since directly correlated to the actual seismic behavior ofbuildings and can be very useful for validation of the others models
ü Vulnerability is dependent on the local seismic culture and the availablematerials in the area, so the extrapolation of empirical fragility functionsfor traditional masonry buildings to other areas can be questionable.
FRAGILITY CURVES: how are they obtained?
q Empirical / Observational – Examples for RC structures
REF: Del Gaudio et al. (2017) Empirical fragility curves from damage data on RC buildings after the 2009 L’Aquila earthquake, Bull Earthqyake Eng 15: 1425-1450
Reference stock: L’AQUILA 2009 earthquake; derivation of fragility curves for sub-typologies
FRAGILITY CURVES: how are they obtained?
q Empirical / Observational – Examples for URM structures
REF: Rosti et al. (2019) Derivazione di curve di fragilità empiriche per edifici residenziali in muratura, ANIDIS Conference, Ascoli Piceno 2019.
Reference stock: L’AQUILA 2009 earthquake; derivation of fragility curves for sub-typologies
FRAGILITY CURVES: how are they obtained?
q Empirical / Observational – some critical issues
0
1
2
3
4
5
4 5 6 7 8 9 10 11 12
µD
I (MCS)
INTENSITY MEASURE
PROBA
BILITY
Issues related to the actual completeness of data for low intensity level of the seismic inputNOT ALL BUILDINGS ARE SISTEMATICALLY
SURVEYED IN THE AREA FAR TO THE EPICENTER
!!
HOMOGENEITY OF THE BUILDING STOCK
Sometimes different trends varying the intensity can be
associated to differentbuildings features
!!
FRAGILITY CURVES: how are they obtained?
q Expertise-based / Heuristic – Expert elicitationü Expert elicitation is used to assess vulnerability of building types, if no data is available and structural analysis is not feasible; one or
more experts can offer an opinion on the level of demand at which damage is likely to occur.ü To process expert judgments the Delphi method (Dalkey, 1969) or the Cooke’s method (Cooke, 1991) can be used.
Collapse fragility estimates obtained using expert elicitation process.
(Jaiswal et al 2013)
Experts responding to target questions at the workshop
Lisbon workshop, September 23, 2012 (Jaiswal et al 2013)Organized by U.S. Geological Survey’s Prompt Assessment for Global Earthquakes Response (PAGER) and Global Earthquake Model (GEM) -
Expert solicitation to develop DPM for 20 building classes, after checking the reliability of experts by seed questions
FRAGILITY CURVES: how are they obtained?
q Mechanical-based
§ Analytical simplified
§ Numerical by nonlinear static analyses
§ Numerical by nonlinear dynamic analyses
archetype buildings are identified and modelled in detail; dispersion of
parameters are related to the whole building stock and not to the
uncertainties of the single building
key features of the building class (structural system, geometry, material
properties) are quantified (median values, dispersion)
Incr
esin
gco
mpu
tatio
nale
ffort
Incr
esin
gam
ount
of d
ata
q BASED ON INSTRUMENTAL INTENSITY MEASURES OF THE SEISMIC INPUT
q THEY ALLOW A HUGE FLEXIBILITY IN SELECTING THE REFERENCE INTENSITY MEASURE TO BE ADOPTED
The macroseismic model starts from the original proposal of Lagomarsino and
Giovinazzi (2006)
REF:Lagomarsino S., Giovinazzi S. (2006) Macroseismic and mechanical models for the vulnerability and damage assessment
of current buildings. Bull Earthquake Eng, 4(4): 415-443
MACROSEISMIC MODEL Then it has further developed by the research group of theUniversity of Genova (UNIGE - S.Lagomarsino,S.Cattari &D.Ottonelli) in recent years through …
a robust calibration on the observed damagecollected after many earthquakes in Italy, available fromthe database Da.D.O. developed by the ItalianDepartment of Civil Protection (DPC) (Dolce et al. 2017).
GRADE 1:Negligible to slight damage
GRADE 2:Moderate damage
GRADE 3:Substantial to heavy damage
GRADE 4:Very heavy damage
GRADE 5:Destruction
D.A.D.O. Database DaDO database: more than 300000 buildings surveyed after 9 different earthquakes in Italy since Friuli 1976
REF: Dolce M., Speranza E., Giordano F., Borzi B., Bocchi F., Conte C., Di Meo A.,Faravelli M., Pascale V. (2019)Observed damage database of past Italian earthqyakes:the Da.D.O. Webgis. Bollettino di Geofisica Teorica e Applicata 60 (2) 141-164.
D.A.D.O. Database DaDO database: more than 300000 buildings surveyed after 9 different earthquakes in Italy since Friuli 1976
DAMAGE SECTION
TYPOLOGICAL SECTION with valuable information also on vulnerability
MACROSEISMIC MODEL – Basics of the original proposal of Lagomarsino & Giovinazzi 2006
q Classifiable as Expert-based q It is directly derived from the European Macroseismic Scale (Grunthal 1998 ), which defines six
vulnerability classes (named from A to F) and various building types (seven of them related to masonry buildings).
GRADE 1:Negligible to slight damage
GRADE 2:Moderate damage
GRADE 3:Substantial to heavy damage
GRADE 4:Very heavy damage
GRADE 5:Destruction
EMS 98
There isn’t a directcorrespondance between a specific structural typology
& a vulnerability class
MACROSEISMIC MODEL – Basics of the original proposal of Lagomarsino & Giovinazzi 2006
q Classifiable as Expert-based q It is directly derived from the European Macroseismic Scale (Grunthal 1998 ), q If a building class is considered, the linguistic definitions of EMS98 may be translated in quantitative terms,
by the fuzzy set theory, and an incomplete Damage states Probability Matrix (DPM) is obtained. The completion is made by using the binomial probability distribution.
EMS 98
μD mean damage ofdistribution
0
0,2
0,4
0,6
0,8
1
0 10 20 30 40 50 60 70 80 90 100
FewManyMost
0
0,2
0,4
0,6
0,8
1
0 10 20 30 40 50 60 70 80 90 100
%DamagedBuilding
FewManyMost
Fuzzy Set Theory
FUZZY SET THEORY BINOMIAL PROBABILITY DISTRIBUTION
translated completed
MACROSEISMIC MODEL – Basics of the original proposal of Lagomarsino & Giovinazzi 2006
q Classifiable as Expert-based q It is directly derived from the European Macroseismic Scale (Grunthal 1998 ), q If a building class is considered, the linguistic definitions of EMS98 may be translated in quantitative terms,
by the fuzzy set theory, and an incomplete (DPM) is obtained.
q The vulnerability is synthetically expressed by a vulnerability curve (Bernardini et al. 2011), which givesthe mean damage μD as a function of the macroseismic intensity I
!! = 2.5 + 3 !"#ℎ! + 6.25! − 12.7
! 0 ≤ !! ≤ 5 !! = 2.5 + 3 !"#ℎ! + 6.25! − 12.7
! 0 ≤ !! ≤ 5
The curve is defined by two parameters representative of the seismic behavior of a group of homogeneous buildings: the
vulnerability index V and the ductility index Q
LS1
LS2
LS3
LS4
LS5
0
1
2
3
4
5
3 4 5 6 7 8 9 10 11 12
µD
MACROSEISMIC INTENSITY
A - V=0.88 B - V=0.72 C - V=0.56 D - V=0.40 E - V=0.24 F - V=0.08
MACROSEISMIC MODEL – Basics of the original proposal of Lagomarsino & Giovinazzi 2006
q Classifiable as Expert-based q It is directly derived from the European Macroseismic Scale (Grunthal 1998 ), q If a building class is considered, the linguistic definitions of EMS98 may be translated in quantitative terms,
by the fuzzy set theory, and an incomplete (DPM) is obtained.
q The vulnerability is synthetically expressed by a vulnerability curve (Bernardini et al. 2011), which givesthe mean damage μD as a function of the macroseismic intensity I
!! = 2.5 + 3 !"#ℎ! + 6.25! − 12.7
! 0 ≤ !! ≤ 5 !! = 2.5 + 3 !"#ℎ! + 6.25! − 12.7
! 0 ≤ !! ≤ 5
The curve is defined by two parameters representative of the seismic behavior of a group of homogeneous buildings: the
vulnerability index V and the ductility index Q
Class A B C D E V 0.9 0.74 0.58 0.42 0.26
Q assumed costant and equal to 2.3 for residential buildings
LS1
LS2
LS3
LS4
LS5
0
1
2
3
4
5
3 4 5 6 7 8 9 10 11 12
µD
MACROSEISMIC INTENSITY
A - V=0.88 B - V=0.72 C - V=0.56 D - V=0.40 E - V=0.24 F - V=0.08
MACROSEISMIC MODEL – Recent developments made by UNIGE
STATISTICAL REPROCESSING OF DATA AND DEFINITION OF DPM
FIRST STEP: Conversion of damage data of AeDES formsinto a DAMAGE LEVEL compatible with that defined at
global scale according to the EMS98
Different proposals ….
q Rota et al. 2008q Pasquale and Goretti 2001q D.A.D.O proposal by DPCq Proposal by UNIGE within the ReLUIS research
DPM for I=6.5
I N° Nuova Proposta Rota, 2008 Goretti-Di Pasquale
(medio) Angeletti, 1982 Goretti-Di Pasquale
(medio) Goretti, 2001
5
5803
5.5
9820
6
1981
8
6.5
3930
7
3284
7.5
8172
8 966
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
Proposal byUNIGE Rotaetal2008 DiPasqualeeGoretti,2001
REF:LagomarsinoS.,CattariS.,Ottonelli D.(2020)Macroseismic fragility curves forItalian residential URMbuildings calibratedbyobserved damage ,Bulletin ofEarthquake Engineering,tobesubmitted.
MACROSEISMIC MODEL and EMPIRICAL DATA1. Conversion of empirical damage data collected with a DAMAGE SURVEY FORM to a DAMAGE LEVEL
DAMAGE TO EACH ELEMENT• Multiple choice possibility• Combination rule function of the
extension of damage: ∑ei ≤ 1• With i each structural or non-
structural element
ESTIMATE OF THE GLOBAL DAMAGE
LEVEL
?DIFFERENT APPROACH
WEIGHTED AVERAGE among walls, floors and roof – according toDAMAGE , EXTENSION and WEIGHT of thedifferent elements
Rota et al. 2008
PEAKS OF DAMAGE among walls, floors and roof
DAMAGE and EXTENSION of the walls
Dolce et al. 2017
AeDES Damage DL
Nullo 0
D1 < 1/3 1
1/3 < D1 < 2/3 1
D1 > 2/3 1
D2 - D3 < 1/3 2
1/3 < D2 - D3 < 2/3 3
D2 – D3> 2/3 3
D4 – D5 < 1/3 4
1/3 < D4 – D5 < 2/3 4
D4 – D5 > 2/3 5
Full survey Partial survey
Walls 0.6 0.8Floors 0.2 0Roof 0.2 0.2
Criteria from Umbria and MarcheEarthquake, 1997
AeDES Damage di
D0 0D1 1
D2 – D3 3D4 – D5 5
Extension ei
> 2/3 5/61/3 - 2/3 1/2
< 1/3 1/6
Lagomarsino et al. 2020
MACROSEISMIC MODEL and EMPIRICAL DATA2. Application of the different criteria and results in terms of DPM
I N° Nuova Proposta Rota, 2008 Goretti-Di Pasquale
(medio) Angeletti, 1982 Goretti-Di Pasquale
(medio) Goretti, 2001
5
5803
5.5
9820
6
1981
8
6.5
3930
7
3284
7.5
8172
8 966
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
I=6.5 I=7.5
I N° Nuova Proposta Rota, 2008 Goretti-Di Pasquale
(medio) Angeletti, 1982 Goretti-Di Pasquale
(medio) Goretti, 2001
5
5803
5.5
9820
6
1981
8
6.5
3930
7
3284
7.
5
8172
8 966
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
WEIGHTED AVERAGE
I N° Nuova Proposta Rota, 2008 Goretti-Di Pasquale
(medio) Angeletti, 1982 Goretti-Di Pasquale
(medio) Goretti, 2001
5
5803
5.5
9820
6
1981
8
6.5
3930
7
3284
7.5
8172
8 966
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
PEAK OF DAMAGE
I N° Nuova Proposta Rota, 2008 Goretti-Di Pasquale
(medio) Angeletti, 1982 Goretti-Di Pasquale
(medio) Goretti, 2001
5
5803
5.5
9820
6
1981
8
6.
5
3930
7
3284
7.5
8172
8 966
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"0"
0.5"
1"
0" 1"2" 3"4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"0"
0.5"
1"
0" 1" 2" 3" 4" 5"
WEIGHTED AVERAGE among walls, floors and roof –
according to DAMAGE , EXTENSION and WEIGHT of the
different elements
PEAKS OF DAMAGE among walls, floors
and roof
MACROSEISMIC MODEL – Recent developments made by UNIGE
0
0.5
1
1.5
2
2.5
3
4 5 6 7 8 9 10
μD
I
DannoUniGE
Emilia2012 L'Aquila2009 Pollino1998Marche1997 Irpinia1980 Friuli1976
0
0.5
1
1.5
2
2.5
3
4 5 6 7 8 9 10
μD
I
DannoDado
Emilia2012 L'Aquila2009 Pollino1998Marche1997 Irpinia1980 Friuli1976
0
0.5
1
1.5
2
2.5
3
4 5 6 7 8 9 10
μD
I
DannoUniGE
Emilia2012 L'Aquila2009 Pollino1998Marche1997 Irpinia1980 Friuli1976
UNIGE Proposal for conversion DADO Proposal for conversion
THE TREND OF INCREASING MEAN DAMAGE WITH INCREASING INTENSITIES IS CONFIRMED , ALSO VARYING
THE APPROACH ADOPTED FOR THE CONVERSION OF AEDES FORM DATA
DATA from Irpinia 1980 and L’Aquila 2009 earthquakesare those more robust for the calibration aims (for number
of data and completeness of area surveyed)
EventoSismico N° EdificiIniziale
N° EdificiconIntensità
N° EdificiinMuratura
Friuli 1976 41852 41852 29641Irpinia1980 38079 33220 26335
Umbria-Marche1997 48525 34873 29512Pollino1998 17442 16689 13887L’Aquila2009 74049 73793 51438Emilia 2012 22554 22489 18194
MACROSEISMIC MODEL – Recent developments made by UNIGE
EVEN THE APPROPRIATENESS OF USING A BINOMIAL DISTRIBUTION IS CONFIRMED
0
0.1
0.2
0.3
0.4
0.5
0 1 2 3 4 5
P
DL
Irpinia80Muratura-I=6.5
BINOMIALE
UNIGE
DADO
0
0.1
0.2
0.3
0.4
0.5
0 1 2 3 4 5
P
DL
Irpinia80Muratura-I=8
BINOMIALE
UNIGE
DADO
MACROSEISMIC MODEL – Recent developments made by UNIGE
CONVERSELY, DATA RELATED TO URM BUILDINGS HIGHLIGHT AS THE Q FACTOR IS NOT CONSTANT VARYING THE CLASSES
0
0.5
1
1.5
2
2.5
3
0 2 4 6 8 10 12
<19-Low
tu3 Q=2.3 Q=3.8
μD
0
0.5
1
1.5
2
2.5
3
0 2 4 6 8 10 12
19-45-Low
tu2 Q=2.3 Q=2.6
0
0.5
1
1.5
2
2.5
3
0 2 4 6 8 10 12
>1981-Medium
tu5 Q=2.3 Q=1.9
L’AQUILA 2009
0
0.5
1
1.5
2
2.5
3
0 2 4 6 8 10 12
Low<1919-Q=3.8 Low19-45-Q=2.6
Low19-61-Q=2.7
μD
SUCHARESULTISCONFIRMEDALSOBYTHEIRPINIA1980DATABASE
In the new proposal in the calibration of the model also the Q factor isconsidered as a free variable
INCREASINGTHEAGETHEQ PARAMETERDECREASES0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
<1900 1900-43 44-61 62-81
Q-In
dicedidu+lità
Classidietà
MEDIUMIRPINIAMEDIUMAQUILALOWIRPINIALOWAQUILA
MACROSEISMIC MODEL – Recent developments made by UNIGE
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura19-45[Num.6688]
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura<1919[Num.29598]
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura46-61[Num.3913]
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura62-81[Num.6267]
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura82-2001[Num.2952]
00.10.20.30.40.50.60.70.80.91
0 1 2 3 4 5
P
DL
AquilaMuratura>2001[Num.780]
μD = 2.06 μD = 2.26 μD = 1.79
μD = 0.32 μD = 0.24 μD = 0.86
DPM FOR L’AQUILA 2009 EARTHQUAKE VARYING THE AGE
EVIDENCES FROM L’AQUILA EARTHQUAKE ON THE CHANGES OF URM BUILDINGS BEHAVIOUR
MACROSEISMIC MODEL – Recent developments made by UNIGE
DEVELOPMENT OF THE MODEL ACCOUNTING FOR THE EVIDENCES FROM REAL DAMAGE DATA
!! = 2.5 + 3 !"#ℎ! + 6.25! − 12.7
! 0 ≤ !! ≤ 5 !" = 2.5 '1 + *+,ℎ./ +6,252− 10.8 − 66 78
Lagomarsino&Giovinazzi 2006 Newproposal withQ=0.9+2.8V
0
1
2
3
4
5
0 1 2 3 4 5 6 7 8 9 10
μD
I
Irpinia1980 Friuli1976 Pollino98 V(2006) V(2011)
V(2006mod) Emilia2012 L'Aquila2009 Marche1997
• V(2006): Lagomarsino e Giovinazzi 2006;• V(2011): Bernardini et al 2011;• V(2006 mod): new proposal
y=2.8066x+0.9135
0
0.5
1
1.5
2
2.5
3
3.5
4
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Q-Indicedidu+
lità
V-Indicedidu+lità
L’Aquila2009
Q =2.8V+0.9
MACROSEISMIC MODEL – Recent developments made by UNIGE
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
4 6 8 10 12
Dann
omed
io
Intensità
A-L&G2006B-L&G2006C-L&G2006D-L&G2006A-new1B-new1C-new1D-new1
Class A B C DV 0.99 0.80 0.61 0.42
!" = 2.5 1 + )*+ℎ - + 3.450 − 11.72.80 + 0.9
New proposal
Updating of reference values of EMS98 classes in order to obtainresults more consistent with data
from observed damage
MACROSEISMIC MODEL – Recent developments made by UNIGE USE OF DATA AVAILABLE ON SUB-CLASSES TO DEFINE PROPER BEHAVIOUR MODIFIERS IN ORDER
TO DEFINE THE EXPECTED CHANGES IN THE CORRESPONDING VULNERABILITY CURVES
0.0
0.2
0.4
0.6
0.8
1.0
1.2
<1900 1900-43 44-61 62-81
V-In
dicedivulnerab
ilità
Classidietà
MEDIUMIRPINIAMEDIUMAQUILALOWIRPINIALOWAQUILA
AGE
REFERENCE MEAN VALUE
OF THE WHOLE CLASS
MODIFIERS
IF > 0 IT INCREASES THE
VULNERABILITY
IF <= 0IT DECREASES THE
VULNERABILITY
MACROSEISMIC MODEL – Recent developments made by UNIGE
USE OF DATA AVAILABLE ON SUB-CLASSES TO DEFINE PROPER BEHAVIOUR MODIFIERS IN ORDER TO DEFINE THE EXPECTED CHANGES IN THE CORRESPONDING VULNERABILITY CURVES
0.0
0.2
0.4
0.6
0.8
1.0
1.2
<1900 1900-43 44-61 62-81
V-In
dicedivulnerab
ilità
Classidietà
MEDIUMIRPINIAMEDIUMAQUILALOWIRPINIALOWAQUILA
MASONRY TYPOLOGYRC= Regular Cut masonry // UC=Uncut masonryQUALITY OF STRUCTURAL DETAILSHQD=High Quality Detailed (tie rods and ring beams) // LQD=Low Quality Detailed (no tie rods and ring beams)DIAPHRAGMS TYPOLOGYV=Vauls // F=Flexible Floor // R=Semi-rigid Floor // RC=Rigid Floor
TAXONOMY of reference (Lagomarsino and Cattari 2013)
AGE
CALIBRATION AND COMBINATION OF BEHAVIOUR MODIFIERS
Independent processCascade process
RC= Regular Cut masonry // UC=Uncut masonryHQD=High Quality Detailed (tie rods and ring beams) // LQD=Low Quality Detailed (no tie rods and ring beams)
V=Vauls // F=Flexible Floor // R=Semi-rigid Floor // RC=Rigid Floor
TAXONOMY of reference (Lagomarsino and Cattari 2013)
GROUPING THE DATA ON SPECIFIC SUB-CLASSES ACCORDING TO ….
CALIBRATION AND COMBINATION OF BEHAVIOUR MODIFIERS
Indipendent Process
V_Global Abruzzo= 0.692 [40781] Buildings] V_Regular Cut Abruzzo = 0.426 [12369] Buildings]V_Uncut Abruzzo = 0.814 [25763] Buildings]
Vulnerability Index
0
1
2
3
4
5
5 6 7 8 9 10 11 12
µ D
Macroseismc Intensity
GlobalRegularUncut
Decreases vulnerabilityΔV<O
Increases vulnerabilityΔV>O
0
1
2
3
4
5
5 6 7 8 9 10 11 12
µ D
Macroseismc Intensity
GlobalHQDLQD
Decreasesvulnerability ΔV<O
Increases vulnerabilityΔV>O
V_Global Abruzzo= 0.692 [40781] Buildings] V_HQD Abruzzo = 0.594 [13749] Buildings]V_LQD Abruzzo = 0.786 [24200] Buildings]
Vulnerability Index
ROLE OF MASONRY TYPE ROLE OF STRUCTURAL DETAILS
CALIBRATION AND COMBINATION OF BEHAVIOUR MODIFIERS
0
1
2
3
4
5
5 6 7 8 9 10 11 12
µ D
Macroseismc Intensity
GlobalFlexibleVaultRigidRigid - RC
Indipendent Process
V_Global Abruzzo= 0.692 [40781] Buildings] V_vault= 0.881 [3864] Buildings]V_rigid = 0.865 [5660] Buildings]V_rigid = 0.631 [9442] Buildings]V_rigid = 0.444 [8300] Buildings]
Vulnerability Index
Decreases vulnerabilityΔV<O
Increases vulnerabilityΔV>O
ROLE OF DIAPHRAGMS
SEISMIC BEHAVIOUR MODIFIERS
Type of masonry
LQD: absence of aseismic devices
HQD: presence of aseismic devices
Modifiers in DVm
-0.25
-0.20
-0.15
-0.10
-0.05
0.00
0.05
Regular Cut Uncut Regular Cut LQD Regular Cut HQD Uncut LQD Uncut HQD
Macroseismic Modifiers - PRE 1919
Vi decreases –Vulnerability decreases
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
<1919-UC-LQD <1919-UC-HQD <1919-RC-LQD <1919-RC-HQD-V
Macroseismic Modifiers
Cascade process SRSS SUM
Modifiers in DVm
Costruction period < 1919 0.132
Masonry quality - RC -0.180
Masonry quality - UC 0.127
Detail quality - LQD 0.112
Detail quality - HQD -0.064
FROM recent developments made by UNIGE: SRSS RULE MORE APPROPRIATE!!
MACROSEISMIC MODEL – from vulnerability curves to fragility curves
q Firstly, it is necessary to define a reference MEAN DAMAGE VALUE to be associated to each DAMAGE LEVEL
LS1
LS2
LS3
LS4
LS5
0
1
2
3
4
5
3 4 5 6 7 8 9 10 11 12
µD
MACROSEISMIC INTENSITY
A - V=0.88 B - V=0.72 C - V=0.56 D - V=0.40 E - V=0.24 F - V=0.08
µD,DL1
µD,DL2
µD,DL5
0
0,1
0,2
0,3
0,4
0 1 2 3 4 5
µD,DL2 = 1.57
µD,DLk = 0.93k − 0.29
Linearregression fromvalues obtained fromthebinomial distribution
MACROSEISMIC MODEL – from vulnerability curves to fragility curves
q Firstly, it is necessary to define a reference MEAND DAMAGE VALUE to be associated to each DAMAGE LEVELq Then, it is possible computing the fragility curve in terms of Intensity by assessing the I value that produces the
attainment of DLk
𝐼"OP = 11.7 − 3.45𝑉 + 0.9 + 2.8𝑉 𝑎𝑡𝑎𝑛ℎ 0.4𝜇"OP − 1 µD,DLk = 0.93k − 0.29
0
0,2
0,4
0,6
0,8
1
0 2 4 6 8 10 12
Prob
abili
tà
Intensità
Classe A
D1D2D3D4D5
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
0 2 4 6 8 10 12Pr
obab
ilità
Intensità
Danno D2
Classe A
Classe B
Classe C
Classe D
ForaGIVENvulnerability class varying theDL ForaGIVENDLvarying thevulnerability class
NOTWELLREPRESENTEDBYTHECUMULATIVELOGNORMALFUNCTION…
MACROSEISMIC MODEL – from vulnerability curves to fragility curves
q Firstly, it is necessary to define a reference MEAND DAMAGE VALUE to be associated to each DAMAGE LEVELq Then, it is possible computing the fragility curve in terms of Intensity by assessing the generic I value that
produces the attainment of DLkq Finally, it is necessary to introduce a proper Intensity – PGA correlation law in order to define the fragility curve in
terms of a instrumental intensity measure
Comparison between some I-PGA Correlation law available in literature and that calibrated by UNIGE on basis of shakemap data from L’Aquila 2009 earthquake
0,0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1,0
4 5 6 7 8 9 10 11 12
PGA
[g]
Intensità Macrosismica
EDIFICI DBMargottini (1992)Murphy and O'Brien (1977)Faccioli e Cauzzi (2006)Faenza e Michelini (2010)L'Aquila 50%L'Aquila 16%L'Aquila 84% CorrelazioneI-PGA c1 c2
Margottini et al. (1992) 0.0430 1.66Murphy and O'Brien (1977) 0.0322 1.78Faccioli e Cauzzi (2006) 0.0464 1.67Faenza e Michelini (2010) 0.0197 2.44
CorrelazioneI-PGA c1 c2daShakeMapL'Aquila(mediana) 0.05 1.66daShakeMapL'Aquila(16%) 0.02 1.82daShakeMapL'Aquila(84%) 0.13 1.48
UNIGEproposal
Literature proposals
MACROSEISMIC MODEL – from vulnerability curves to fragility curves
q Firstly, it is necessary to define a reference MEAND DAMAGE VALUE to be associated to each DAMAGE LEVELq Then, it is possible computing the fragility curve in terms of Intensity by assessing the generic I value that
produces the attainment of DLkq Finally, it is necessary to introduce a proper Intensity – PGA correlation law in order to define the fragility curve
in terms of a instrumental intensity measure
0
0,2
0,4
0,6
0,8
1
0 0,2 0,4 0,6 0,8
Prob
abilit
à
PGA [g]
Classe A
D1 - log D2 - log D3 - log D4 - log D5 - logD1 D2 D3 D4 D5
The fragility curve in PGA is wellfitted by the lognormal cumulative
function !
!!" !" = ! ! > !!" !" = ! !"!" < !" = Φ!"# !"
!"!"!!"
!"#$% = '(')(+,-./) = '(')1.3.4.)/56(7.86).95):;:<= 7.41%.(.79
MACROSEISMIC MODEL –fragility curves developed by UNIGE
q fragility curve in terms of a instrumental intensity measure
WHAT SOURCES OF UNCERTAINTY ARE CONSIDERED?
0
0,2
0,4
0,6
0,8
1
0 0,2 0,4 0,6 0,8
Prob
abilit
à
PGA [g]
Classe A
D1 - log D2 - log D3 - log D4 - log D5 - logD1 D2 D3 D4 D5
βLS
IMLS
From the fittingUncertainty on the V value - variability at
large scale of vulnerability classes
L’Aquila 16° quantile – Irpinia 84° quantile
!"# = !%% &, (%, ) + !+% )
MACROSEISMIC MODEL –fragility curves developed by UNIGE
HOW WE CAN PASS FROM THE FRAGILITY CURVE OF THE EMS98 VULNERABILITY CLASSES TO OTHER SUB-CLASSES
(à TARGETED TO OUR INVENTORY & OUR AVAILABLE DATA)?
0
0,2
0,4
0,6
0,8
1
<1919 19-45 46-61 62-81 >1981
V
Classi di età
LowMediumHigh
LOWClassi età VEMPIRICI A B C D< 1919 0.952 80 201919 - 1945 0.847 25 751946 - 1961 0.705 50 501962 - 1981 0.550 70 30> 1981 0.420 100
MEDIUMClassi età VEMPIRICI A B C D< 1919 0.914 60 401919 - 1945 0.781 90 101946 - 1961 0.743 70 301962 - 1981 0.648 20 80> 1981 0.496 40 60
Class A B C DV 0.99 0.80 0.61 0.42
BY ASSIGNING PROPER% A - % B - % C - % D
SUB-CLASS
NEW MACROSEISMIC MODEL –VALIDATION
ReLUIS-DPC Project: Italian seismic risk map
For the aim of validation and within the contextof ReLUIS-DPC project addressed to developingItalian seismic risk map the fragility curves have
been implemented in the IRMA Platform
ReferenceefiguradiIRMA
DL1– Simulated bythemacroseismic model
ScenarioofL’Aquila2009earthquake – Validation madebythePlatformIRMADL1– real datafrom
DaDO
NEW MACROSEISMIC MODEL –VALIDATION
DL3– Simulated bythemacroseismic model
ScenarioofL’Aquila2009earthquake- Validation madebythePlatformIRMADL3 – real datafrom
DaDO
NEW MACROSEISMIC MODEL –VALIDATION
DL4– Simulated bythemacroseismic model
DL4– real datafromDaDO
NEW MACROSEISMIC MODEL –VALIDATION
ScenarioofL’Aquila2009earthquake- Validation madebythePlatformIRMA
Barisciano
Sant’EusanioForconese
Shakemap ofL’Aquila2009
NEW MACROSEISMIC MODEL –VALIDATION
COMPARISON IN TERMS OF DPM FOR VARIOUS MUNICIPALITIES WITH DIFFERENT EPICENTRAL DISTANCE
0
0.1
0.2
0.3
0.4
0.5
0 1 2 3 4 5
Probabilità
Danno
L'Aquila
DaDO
IRMA
DANNO0 1 2 3 4 5 TOT.
5276 1686 1639 970 628 490 106891742 1712 2180 2174 1374 254 9436
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5
Probabilità
Danno
Barisciano
DaDOIRMA
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5
Probabilità
Danno
Sant'Eusanio
DaDO
IRMA
DANNO0 1 2 3 4 5 TOT. 64 61 71 39 25 40 30042 45 56 54 31 5 234
DANNO0 1 2 3 4 5 TOT. 539 231 204 74 31 18 1097285 245 226 161 62 7 986
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0 1 2 3 4 5
Probabilità
Danno
PoggioPicenze
DaDO
DANNO
0 1 2 3 4 5 TOT.
128 63 63 44 14 28 34080 74 76 60 27 3 320
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 1 2 3 4 5
Probabilità
Danno
Sulmona
DaDO
IRMA
DANNO0 1 2 3 4 5 TOT. 778 149 64 12 4 0 10071713 490 116 21 1 0 2341
sulmona
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)
m*
h*
m*
h*
d
m*
h* zi
mi, ψι#
XYSd
Sa
Du,X
Au,X
4π2/Ty,X2#
q Possibility of taking into account the variousparameters that determine the structuralresponse
q Need to define a statistical model for randomvariables
q Definition of the LIMIT STATES on the capacitycurve and the corresponding damping values
Corrective factors
Κ1,Κ2,Κ3 , Κ4
for the evaluationof the yieldingacceleration
Κ5 , Κ6 for the evaluationof the period
For the random variables
84%
0"
0,01"
0,02"
0,03"
0,04"
0,05"
0,06"
0,07"
40" 45" 50" 55" 60" 65" 70" 75" 80" 85" 90"
16%
50%
Mechanical parameters and loadsτk,x shear strengthGx,i shear modulusγi specific weight
ΔS,LS4 ΔF,LS4
drift limit values for the shear and flexural response
κx,i spandrel contribution on the masses
qi - ζx,i load and orientation of floors
Geometrical features
hi inter-storey heightαx,i ratio of resistant wall
over the gross area A
Modal shape
φx,i i-th component of theassumed modal shape
Failure mode
εx weight assigned to thein-plane SSWPmechanism
REF:Lagomarsino, S., Cattari S. (2014). Fragility functions of masonry buildings (Chapter 5), pp.111-156. In SYNER-G: Typology Definition and Fragility Functions for Physical Elements at Seismic Risk: Elements at Seismic Risk, Geotechnical, Geological and Earthquake Engineering 27 pp. 420
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
ACCELERATION ULTIMATE DISPLACEMENT CAPACITY
PERIOD
SSW
P
WSS
P
KSTIFFNESS - Corrective factors to account for the flexuralcontribution in piers and the role of spandrels
KSTRENGTH - Corrective factors to account for the irregularity and the role of spandrels
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
Sd
Sa
Du,X
Au,X
4π2/Ty,X2#
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
AC
CE
LER
ATIO
N
DISPLACEMENT
WSS
P EF
SS
WP
OV
ER
ALL
BA
SE
SH
EA
R
Presence of r.c. tie beams
Presence of steel tie – rods
Spandrels not coupled to other tensile resistant element
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
The evaluation of these variables requires:
• the definition of a limited number of mechanical and geometrical parameters
• the assumption of a fundamental modal shape
• the attribution of specific correction factors, aimed to take into account the effects related to the comprehensive set of constructive and morphologicaldetails
e.g. : the presence of tie-rods or r.c. ring beams coupledto the spandrels
PERIOD
KSTIFFNESS - Corrective factors to account for the flexuralcontribution in piers and the role of spandrels
ACCELERATION
KSTRENGTH - Corrective factors to account for the irregularity and the role of spandrels
ULTIMATE DISPLACEMENT CAPACITY
SSW
P
WSS
P
aims to consider the flexuralcontribution in piersK5
K 5 = 1+ 11.2
GE
hb
bp
⎛
⎝⎜⎞
⎠⎟
2⎡
⎣⎢⎢
⎤
⎦⎥⎥
−1
PERIOD
KSTIFFNESS - Corrective factors to account for the flexuralcontribution in piers and the role of spandrels
effects on the stiffness related to the spandrelinfluence on the boundary conditions on piersK6
Correction factor EF WSSP SSWPK5 0.4 ÷ 0.8 0.4 ÷ 0.8 0.6 ÷ 0.8K6 0.6 ÷ 1 0.3 ÷ 0.7 1
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
AC
CE
LER
ATIO
N
DISPLACEMENT
WSS
P EF
SS
WP
KSTIFFNESS
ACCELERATION
KSTRENGTH - Corrective factors to account for the irregularity and the role of spandrels
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
AC
CE
LER
ATIO
N
DISPLACEMENT
WSS
P EF
SS
WP
KSTRENGTH
modifies the strength as a function of the main prevailingfailure mode expected at scale of masonry piers.K1
accounts for the influence of thenon homogeneous size of themasonry piers.
K2 K 2 = 1− 0.2Nmxl Axl , j
2
j∑
Axl
2≥ 0.8
K 3 = 1+eyldyl Axl
yk − yCl( )2Axl ,k
k∑
≤1.25
accounts for the influence ofgeometric and shape irregularitiesin the plan configuration.
K3
accounts for the effectiveness of spandrels, whichinfluence the global failure mechanism of the building (EF,WSSP and SSWP).
K4
Correction factor EF WSSP SSWP
K1 0.8 ÷ 1.5 0.8 ÷ 1 1 ÷ 1.5
K2 0.8 ÷ 1 1 1
K3 0.75 ÷ 1 0.75 ÷ 1 0.75 ÷ 1
K4 0.6 ÷ 1 0.5 ÷ 0.8 1
SSW
P
WSS
P
• The same formula also applies to LS3
• Dy,X=Au,X(Ty,X/2p)2 is the yielding displacement
• N: number of stories (i=1,..N is the level counter)
• eX allows considering intermediate failure modes, which occurin Equivalent Frame (EF) behavior
ULTIMATE DISPLACEMENT CAPACITY
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
SSW
P
WSS
P
• The same formula also applies to LS3
• Dy,X=Au,X(Ty,X/2p)2 is the yielding displacement
• N: number of stories (i=1,..N is the level counter)
• eX allows considering intermediate failure modes, which occurin Equivalent Frame (EF) behavior
DS (shear) DF (flexural)LS3 0.0025 - 0.004 0.004 - 0.008LS4 0.004 - 0.006 0.008 - 0.012
• Identification of the DL1 and DL2DLS1 =0.7 Dy,XDLS1 =C2 Dy,X
Sd
DEFINITION OF DL
Say
SduSdy
?
? ?
ULTIMATE DISPLACEMENT CAPACITY
§ c2 is a coefficient that varies as a function of the prevailing global failure mode.It is proposed to assume a value for c2 from 1.2 to 2 (Lagomarsino and Cattari2014), varying from the SSWP to the WSSP failure mode
DL1
DL2
DL3
DL4
THE MECHANICAL–BASED APPROACH : The DVB-masonry model proposed in Lagomarsino & Cattari 2014
DL5 ????
• DLS4 is alternatively assumed equal to DS,LS4 or DF,LS4 as afunction of the prevailing failure mode in masonry piers for theexamined direction (if shear or flexural one)
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum
IMLS
IMLS Sa(Sd)
(DLS , ALS)
IMLS Sa(Sd) η(x)
LS
Sa(Sd) 1
TLS
overdamped spectrum elastic spectrum
spectrum normalized
AC
CE
LER
AT
ION
DISPLACEMENT
!"!" =!!"
!!1(!!")!(!!")
!(!!") =10
5 + !!"
REF:Discussion on the reliability of nonlinear static procedures in Marino et al. 2019 (Engineering Structures, in press)
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
𝛽𝐿𝑆 = %𝛽𝐻2 + 𝛽𝐷
2 + 𝛽𝑇2 + 𝛽𝐶
2
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
𝛽𝐿𝑆 = %𝛽𝐻2 + 𝛽𝐷
2 + 𝛽𝑇2 + 𝛽𝐶
2 !"# = %!&2 + !)2 + !*2 + !+2
!! = 0.5 !"# !"!,84 !(!"!") − !"# !"!,16 !(!"!")
λ16" λ"
λ84"
λ(IMLS)
IMLS
IMH,84[λ(IMLS)]
IMH,16[λ(IMLS)]
λ"
IM
IMH,16[λ(IMLS)]
Sa1(Sd)
IMH,84[λ(IMLS)]
IMLS
SPEC
TRAL
AC
CEL
ERAT
ION
SPECTRAL DISPLACEMENT
Influence on the spectral demand of the epistemic uncertainty on the hazard curve
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
𝛽𝐿𝑆 = %𝛽𝐻2 + 𝛽𝐷
2 + 𝛽𝑇2 + 𝛽𝐶
2
!! = 0.5 !"# !"!,84 − !"# !"!,16
CAPACITY DEMAND
For the random variables
84%
0"
0,01"
0,02"
0,03"
0,04"
0,05"
0,06"
0,07"
40" 45" 50" 55" 60" 65" 70" 75" 80" 85" 90"
16%
50% 16 e 84% Sa
T
ItrequirestheevaluationoftheintensitymeasuresIMD,16 andIMD,84 that
correspondtoadisplacementdemandequaltoDLS,onthemediancapacitycurveoftheconsideredclassof
buildings,byusingtheconfidencelevelsresponsespectraSa1,16(Sd)andSa1,84(Sd)
respectively
!! = 0.5 !"# !"!,84 − !"# !"!,16
Sa1,16(Sd)
Sa1(Sd)
Sa1,84(Sd)
PGA=1 SP
ECTR
AL A
CC
ELER
ATIO
N
SPECTRAL DISPLACEMENT
Sa1,16(Sd) Sa1(Sd)
Sa1,84(Sd)
Sa,max=1
SPEC
TRAL
AC
CEL
ERAT
ION
SPECTRAL DISPLACEMENT
IM=PGA IM=Sa(T1)
Influence of the selection of IM
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
𝛽𝐿𝑆 = %𝛽𝐻2 + 𝛽𝐷
2 + 𝛽𝑇2 + 𝛽𝐶
2
!! = 0.5 !"# !"!,84 − !"# !"!,16
DLS DLS,84
LS1
LS2 LS3 LS4
p p p
DLS
p
AC
CE
LER
AT
ION
DISPLACEMENT DLS,16
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
𝛽𝐿𝑆 = %𝛽𝐻2 + 𝛽𝐷
2 + 𝛽𝑇2 + 𝛽𝐶
2
! = !!! !!!!!
xk =(Xk-μk)/σk
CAPACITY DEMAND2N analyses
50% Sa
T
!! = !!!
Through the use of the RESPONSE SURFACE TECNIQUE
THE MECHANICAL –BASED APPROACH: definition of fragility curves
q Classifiable as Analytical method q Structural response described in terms of CAPACITY CURVE (only the GLOBAL IN-PLANE response is considered)q Evaluation of IMDL through the comparison between the CAPACITY CURVE and the SEISMIC INPUT
expressed by the response spectrum q Evaluation of bDL through the ANALYTICAL EVALUATION of all sources of uncertainties
FRAGILITY CURVEofDamage States DAMAGE PROBABILITY
&IMLS
IMLS Sa(Sd)
(DLS , ALS)
IMLS Sa(Sd) η(x)
LS
Sa(Sd) 1
TLS
overdamped spectrum elastic spectrum
spectrum normalized
AC
CE
LER
ATIO
N
DISPLACEMENT
VARIATION OF MASONRY QUALITY
VARIATION OF FLOORS
LQD HQD
VARIATION OF NUMBER OF STOREYS
VARIATION OF QUALITY DETAILS
HQD
APPLICATION of ANALYTICAL MODELS FOR MASONRY BUILDINGS
LQD HQD
SEISMIC BEHAVIOUR MODIFIERS
Type of masonry
LQD: absence of aseismic devices
HQD: presence of aseismic devices
Type of masonry
LQD: absence of aseismic devices
HQD: presence of aseismic devicesRB - Ring BeamsTR – Tie Rods
Modifiers in DPGA
Modifiers in DPGA
-0.06
-0.01
0.04
0.09
0.14
0.19
Regular Cut Uncut Regular Cut LQD Regular Cut HQD Uncut LQD Uncut HQD
Macroseismic Modifiers - PRE 1919
-0.06
-0.01
0.04
0.09
0.14
0.19
Regular Cut Uncut Regular CutLQD
Regular CutHQD TR
Regular CutHQD TB
Uncut LQD Uncut LQDTR
Uncut LQDRB
Mechanical Modifiers - PRE 1919
PGA increases –Vulnerability decreases
PGA increases –Vulnerability decreases
THE MECHANICAL –BASED APPROACH: fragility curves varying the sub-classes
Vault
Flexible
Rigid
Reinforced concrete
Vault
Flexible
Rigid
Reinforced concrete
Vault
Flexible
Rigid
Reinforced concrete
Low Quality Details
High Quality Details (Tie rods)
Low Quality Details (Ring beams)
Uncut masonry
Masonry – < 1919 –Number of storeys 3
DEFINITION of the CLASSES of REFERENCE
Thesame forthecut masonry……
DIAPHRAGMS GROUPED DETAILS GROUPEDMASONRY TYPOLOGY GROUPED
Combination of the fragility curves to obtain the curve representative of the class: MASONRY – < 1919 – NUMBER OF STOREYS 3
DIAPHRAGMS GROUPED
DETAILS GROUPED
MASONRY TYPOLOGY GROUPED
THE MECHANICAL –BASED APPROACH: fragility curves varying the sub-classes
…..Combination of the fragility curves to obtain the curve representative of the class:
MASONRY – < 1919 – NUMBER OF STOREYS 3
Uncut Masonry 86%Regular cut Masonry 14%
Uncut Masonry Low Quality Details 77%High Quality Details 23%
Regular cut Masonry Low Quality Details 62%High Quality Details 38%
Uncut Masonry
LQD
Vault 26%Flexible 34%
Rigid 36%Reinforcedconcrete
4%
HQD
Vault 24%Flexible 24%
Rigid 39%Reinforcedconcrete
13%
Regular cut Masonry
LQD
Vault 20%Flexible 22%
Rigid 47%Reinforcedconcrete
11%
HQD
Vault 13%Flexible 13%
Rigid 34%Reinforcedconcrete
40%
DIAPHRAGMS GROUPED
DETAILS GROUPED
MASONRY TYPOLOGY GROUPED
THE MECHANICAL –BASED APPROACH: fragility curves varying the sub-classes
• Built in 1930• Interstory heigth: 4.00 m• Gross area:520 m2
• T-shaped plan• 3 story building• Rigid floors• Stone masonry
The school of P. Capuzi in Visso monitored by the Italian Department of Civil Protection
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
Dataavailable fromReLUIS Project2017/2018.REF:S.Cattarietal.2019DiscussionondatarecordedbytheItalianstructuralseismicmonitoringnetworkonthreemasonrystructureshitbythe2016-2017CentralItalyearthquake,Proc.ofCOMPDYN2019,Crete24-26June2019.
24/8 26/10 30/10
24/8 26/10 30/10
• Significant damage accumulation effects after the subsequent shocks
The school of P. Capuzi in Visso
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
0
5
10
15
20
0.0 0.5 1.0 1.5 2.0 Sa
[m/s
2 ]
T [s]
24/08/16 26/10/16 30/10/16 NTC
0
5
10
15
20
0.0 0.5 1.0 1.5 2.0
Sa [m
/s2 ]
T [s]
24/08/16 26/10/16 30/10/16 NTC
X directionY direction
NTC = CODE SPECTRUM COMPATIBLE FOR TR=712 years - Soil C
The school of P. Capuzi in Visso
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
Mechanical modelapplied withthespecific parameters ofVissoSchool(interms ofresistant area,masonry typology,interstorey
height,…)
24/08
26/10
24/08
26/10
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
Validation through theDVB-masonry modelmadebyS.Cattari,D.Ottonelli &S.Alfano
24 Agosto
Mud X = 2.986
Mud Y = 2.720
DL0 DL1 DL2 DL3 DL40
0.1
0.2
0.3
0.4
0.5
0.6Dir XDir Y
26 Ottobre
Mud X = 3.110
Mud Y = 3.174
DL0 DL1 DL2 DL3 DL40
0.1
0.2
0.3
0.4
0.5
0.6Dir XDir Y
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
Validation through theDVB-masonry modelmadebyS.Cattari,D.Ottonelli &S.Alfano
24 Agosto
Mud X = 2.986
Mud Y = 2.720
DL0 DL1 DL2 DL3 DL40
0.1
0.2
0.3
0.4
0.5
0.6Dir XDir Y
26 Ottobre
Mud X = 3.110
Mud Y = 3.174
DL0 DL1 DL2 DL3 DL40
0.1
0.2
0.3
0.4
0.5
0.6Dir XDir Y
THE P:CAPUZI SCHOOL – preliminary validation of mechanical model
DL2/DL3 DL3/DL4inY(withDL5inlocal parts)DL2/DL3inX
Validation through theDVB-masonry modelmadebyS.Cattari,D.Ottonelli &S.Alfano
OUTLINE OF THE PRESENTATION
What do theydepend on?
How are theyobtained?
What do theyrepresent?
How can they beused?
Fromvulnerability torisk assessment
Practical issues & Examples of application
Overview & focus on macroseismicand mechanical-based approaches
Relationship with risk analyses &useful GLOSSARY ….
Involved uncertainties & influenceon results
ReLUIS-DPC Project: Italian seismic risk map (2018 - ONGOING)
the IRMA Platform - already used for elaborating in 2018 the first set of Italian seismic risk map – is ongoing to be improved in the MARS
project (Coord. Proff.S.Lagomarsino & A.Masi)
ReferenceefiguradiIRMA
REF. National Risk Assessment (2018) Overview of thepotential major disasters in Italy: seismic, volcanic,tsunami, hydro-geological/hydraulic and extremeweather, droughts and forest fire risks, Presidency ofthe Council of Ministers Italian Civil ProtectionDepartment.REF. Dolce et al. (2019) Seismic risk maps for theItalian territory, XVIII ANIDIS Conference, Ascoli Piceno2019
ReLUIS-DPC Project: Italian seismic risk map (2018 - ONGOING)
EXAMPLES OF MAPS THAT CAN BE PRODUCED (through the implementation of the macroseismic model developed by UNIGE)
Scenario conditioned to 475 years – soil A
DL3 Mean damage – soil A
Scenario unconditioned to 1 yearTAXANOMY:STRUCTURALTYPOLOGY– AGE– HEIGHTCLASS
ReLUIS-DPC Project: Italian seismic risk map (2018 - ONGOING)
Other research groups partecipated to ReLUIS project (from Padua, Naple, Pavia) by defininig fragility curves through different approaches (empirical, hybrid mechanical-based)
REF. Dolce et al. (2019) Seismic risk maps for the Italian territory, XVIII ANIDIS Conference, Ascoli Piceno 2019
The result of maps in terms of damage scenario have been used to assess also the expected LOSSES
It requires theintroduction ofproper correlation laws
EXAMPLESofcorrelation laws between theDAMAGELEVELSand:
CASUALTIES USABILITYSAFEFORUSE/NOTSAFEFORUSE/COLLAPSE
DIRECTECONOMICLOSSRECONSTRUCTIONCOSTS
IRPINIA 1980MESSINA 1908
Loss of life orserious injury requiring hospitalization
ReLUIS-DPC Project: Italian seismic risk map (2018 - ONGOING)
Other research groups partecipated to ReLUIS project (from Padua, Naple, Pavia) by defininig fragility curves through different approaches (empirical, hybrid mechanical-based)
REF. Dolce et al. (2019) Seismic risk maps for the Italian territory, XVIII ANIDIS Conference, Ascoli Piceno 2019
USABILITYSAFEFORUSE/NOTSAFEFORUSE/COLLAPSE
DIRECTECONOMICLOSSRECONSTRUCTIONCOSTS
RESULT FROM THE UNCONDITIONED EVALUATION AT 1 YEAR
COSTinBillions
RESONABLE NUMBERS IF COMPARED WITH THE EARTHQUAKE HYSTORY OF LAST 50 YEARS IN ITALY BUT SIGNIFICANT DISPERSION DUE TO DIFFENCES IN VARIOUS MODELS ADOPTED
RESEARCHONGOINGIN2019WITHINMARS– ReLUIS PROJECT!!!!
CASUALTIES
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
Amatrice, 2016
Accumoli, 2016
Villa Sant’Angelo, L’AQUILA 2009
Amandola, FM, 2016
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
Inte
nsity
Loss of functionality of the urban system
O
D
LS
NC E
Dwellings are compromised
Normal urban function are damaged
Main urban function are interrupted Strategic urban function are interrupted
Strategic emergency function are interrupted
AT URBAN SCALE AT SCALE OF THE SINGLE BUILDING
The Italian Department of Civil Protection developed in recent years specific procedure to support and improve the emergency managment phase as the I.Opà.CLE method addressed
to define Indices for evaluation of the Operational efficiency of Limit Condition Emergency
REF: Dolce M. et al.(2018) Probabilistic assessment of structural operational efficiency in emergency limit conditions: the I.Opà.CLE method, Bull Earthquake Eng 16:3791-3818
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
REF: Dolce M. et al.(2018) Probabilistic assessment of structural operational efficiency in emergency limit conditions: the I.Opà.CLE method, Bull Earthquake Eng 16:3791-3818
THE SISTEM IS EFFICIENT IF ALL COMPONENTS ARE EFFICIENT: THE STRATEGIC FUNCTIONS, THE EMERGENCY AREAS AND THE CONNECTIONS !!
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
REF: Dolce M. et al.(2018) Probabilistic assessment of structural operational efficiency in emergency limit conditions: the I.Opà.CLE method, Bull Earthquake Eng 16:3791-3818
THE SISTEM IS EFFICIENT IF ALL COMPONENTS ARE EFFICIENT: THE STRATEGIC FUNCTIONS, THE EMERGENCY AREAS AND THE CONNECTIONS !!
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
REF: Dolce M. et al.(2018) Probabilistic assessment of structural operational efficiency in emergency limit conditions: the I.Opà.CLE method, Bull Earthquake Eng 16:3791-3818
SPECIFIC FORMS DEVELOPED BY DPC TO PROVIDE RAPID SURVEY AT SCALE OF THE URBAN SYSTEM AND ACQUIRE ESSENTIAL INFORMATION ON STRUCTURES
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
Use of the macroseismic method to define fragility curves and then damage scenario of the buildings stock on the connections to assess their efficiency
E’ necessario associare i DLi al superamento dello Stato Limite di Operatività (inteso come operatività della strada).
C.A.MURATURA
5 6 7 8 9 10 11 12
µD
I(MCS)
P(DLi)
(PGA)
CURVADIVULNERABILITA’ CURVADIFRAGILITA’
The application in this context requires to define the damage threshold NOT COMPATIBLE with the OPERATIONAL EFFICIENCY of CONNECTION
E’ necessario associare i DLi al superamento dello Stato Limite di Operatività (inteso come operatività della strada).
C.A.MURATURA
5 6 7 8 9 10 11 12
µD
I(MCS)
P(DLi)
(PGA)
CURVADIVULNERABILITA’ CURVADIFRAGILITA’
For URM buildingsthe failure associated
to out-of-planemechanisms
For RC buildings the failure associated to
out-of-plane failure of infills
D2/D3D3/D4
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
RESEARCH PROJECT FOUNDED BY THE LIGURIA REGION TO APPLY AND DEVELOP THESE METHODS IN THE SANREMO MUNICIPALITY IN COLLABORATION WITH DICCA
AND DISTAV DEPARTMENTS AND DPC
COMUNE DI SANREMO
USE OF THE MACROSEISMIC MODEL TO SUPPORT THE DESIGN OF SEISMIC EMERGENCY PLAN
TRASCURABILE
MODERATO
SIGNIFICATIVI
SIGNIFICATIVI
CONDITIONED SCENARIO AT 101 YEARS CONDITIONED SCENARIO AT 475 YEARS
Exceedance probability oftheDLassumed as reference fortheNOTcompatible damage
NEGLIBIGLE
MODERATE
HIGH
VERYHIGH
THANK YOU FOR YOUR KIND ATTENTION!
?Serena Cattari
DICCA-DepartmentofCivil,ChemicalandEnvironmentalEngineering