Recent Shaking Table Tests in Portugal: Lessons Learned 2012 earthquake engineering loure… ·...

Recent Shaking Table Tests in Portugal: Lessons Learned

Paulo B. Lourenço

[email protected]/masonry

Institute for Sustainability and Innovation in Structural Engineering

2|Recent Shaking Table Tests in Portugal: Lessons Learned Paulo B. Lourenço

Contents

Recently Completed

• Masonry Structures With Box Behavior

• Masonry Structures Without Box Behavior & Rocking Motion

• Masonry Infills

Just Carried Out & Planned for 2013

• Masonry Infills #2

• Tall timber buildings

• Severe torsional effects

• Dissipative anchors



What is Masonry?



Why is this relevant for mechanics?

Shear testing ofstone joints

-5000

-4000

-3000

-2000

-1000

0

1000

2000

3000

4000

5000

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

-0.4

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

-1.5 -1 -0.5 0 0.5 1 1.5 1ºC

.2º

C.

3º C

.4º

C.

5º C

.6º

C.

7º C

.

Stress-strainrelation

Dilatancy

= -0.179r2 = 0.866

= -0.618r2 = 0.975

= -0.561r2 = 0.985

0.00

0.25

0.50

0.75

1.00

1.25

1.50

-2.00 -1.50 -1.00 -0.50 0.00Normal stress (N/mm2)

She

ar s

tres

s (N

/mm

2 )

P seriesS seriesR seriesLinear (P series)Linear (S series)Linear (R series)Failure

surface



Why is this relevant for mechanics?

Stone walls

Collapse Mechanism and StrengthRegular – tan = 0.4Irregular – tan = 0.3Rubble – tan = 0.2



Linear elastic analysis?“Ut tensio sic vis” or / E = is the elasticity law established by R. Hooke in 1676.The theory is so extensively used that its limitations and deficiencies are often forgotten. This is in opposition with early forms of limit analysis

Cantilever beam according to Galileo (1638) and evolution of the “hypothesis” for the stress distribution at AB

Retaining wall accordingto Coulomb (1773)

B

A

P



Or non-linear analysis?

Structural collapse does not generally coincide with the appearance of the first crack or localized early crushing → elasticity theory is a step back with respect to limit analysis?

Non-linear analysis includes the full loading process, from absence of loading, through behavior under service loading, until collapse → the most advanced form of structural analysis?

Interest growing since 1970’s → it remains a field for specialists due to complexity (knowledge) and costs (time) involved

Possibilities are immense and it is often included in commercial software, but an incorrect use can be very dangerous



The use of non-linear analysis The modern use of non-linear analysis focuses mostly on:

Complex / stringent safety requirement structures (e.g. nuclear plants, dams, bridges…)

A tool to understand in detail results from laboratory tests

Virtual laboratory for parametric studies → Code drafting / Design rules

Existing structures & Earthquake design

Three types of non-linearities are possible:

Material (or physical) non-linearity

F

Geometricalnon-linearity

Contact non-linearity

Steel(“Code”)

Concrete(“Code”)



Existing buildings

Pounding Settlements Vehicles

Earthquake globalcollapse



Non-linear time history analysis

Static non-linear analysis

Linear elastic time history analysis

Modal superposition

Linear static analysis

Sim

plifi

catio

n

Structural Analysis Methods & Earthquake Engineering

Masonry Structures With Box Behavior



Recent test results: Rigid diaphragm Worst case scenario: Embedded ring beam + Unfilled vertical joints

Moderate damage up to 100% of the design earthquake in Lisbon

Ductile failure for 250% of the design earthquake in Lisbon



Earthquake

Lintel

Painel pilar

Ultimate capacity

Linear elastic capacity

“POR” Storey Mechanism



In a force based method, the non-linear reserve capacity must be considered.Displacement based techniques are easy to apply and “standard”

For unreinforced masonry buildings with 2 or more storeys:

EC8:

q = 1.5-2.5 (recommended 1.5)

OPCM 3431:

αu /α1 (OSR) = 1.8

q = q0 x OSR = 3.6

Energy Dissipation Capacity



Continuum Finite Elements (FEM)

FEM is currently a popular approach in structural analysis software, mostly considering isotropic and homogeneous materials. The computational effort is significant if applied at a full building level

Discontinuous Models (FEM, DEM, Limit Analysis)The use of this models is even more complex in terms of computational efforts, making these methods relevant for the study of structural components and research

Structural Component Models: Macro-elements

These seem to be the most appropriate models for design and assessment, with major contributions in Italy

Res

earc

h a

nd

Dev

elo

pm

ent

Des

ign

Recent Analysis Methods



Program Country Code Approach Web adressAEDES Italy Italian FEM and SCM www.aedes.itCMT+L Spain Eurocode FEM www.arktec.com/cmtl.htmFEDRA Norway Eurocode FEM www.runet-software.com/FEDRA.htm

WIN-Statik MurDim+ Sweden ? ? www.strusoft.comPor 2000 Italy Italian SCM www.newsoft-eng.it/Por2000.htm

TQS CAD/Alvest Brazil Brazilian ? www.tqs.com.br/v13/alvest.htmTricalc.13 Spain Eurocode FEM www.arktec.com/new_t13.htmTricalc.17 Spain Spanish FEM www.arktec.com/new_t17.htmWinMason USA USA Storey Mech. www.archonengineering.com/winmason.html

3Muri Italy Italian SCM www.stadata.comANDILWall Italy Italian SCM www.crsoft.it/andilwallMURATS Italy Italian Storey Mech. www.softwareparadiso.it/murats.htmSismur Italy Italian Storey Mech. www.franiac.it/sismur.html

TRAVILOG Italy Italian Storey Mech. www.logical.it/software_travilog.aspxTecnobit Italy Italian Storey Mech. www.tecnobit.info/products/murature.phpCDMaWin Italy Italian FEM and SCM www.stsweb.net/STSWeb/ITA/homepage.htm

There are many commercial softwares available in the market for structural masonry, particularly in Italy. Benchmarking was made in two publications: Azores 1998, Eds. C. Sousa Oliveira et al., (2008) and Marques, R., Lourenço, P.B., Possibilities and comparison of structural component models for the seismic assessment of masonry buildings, Computers and Structures, 89 (21-22), p. 2079-2091 (2011).

Commercial Software



6.50

6.50

1.50 1.00 1.50 1.00 1.50

3.00

2.00

0.30

6.5

0

6.50

1.50 1.00 1.50 1.00 1.50

6.0

0

2.0

01

.00

0.30

6.50

6.50

1.50 1.00 1.50 1.00 1.50

9.00

2.0

01

.00

1.00

0.30

Peso próprio = 6.0 kN/m2 Sobrecarga = 1.5 kN/m2

Peso próprio = 6.0 kN/m2 Sobrecarga = 1.0 kN/m2

Pp = 7.0 kN/m2; Sob = 2.0 kN/m2

Pp = 6.0 kN/m2; Sob = 1.0 kN/m2

Pp = 7.0 kN/m2; Sob = 2.0 kN/m2

Pp = 7.0 kN/m2; Sob = 2.0 kN/m2

Application



1 storey 2 storeys 3 storeys

KO OK for soil type A OK for soil types A and B

Application: Push-over method / OPCM 3431



Application: Elastic design (EC8)

1 storey 2 storeys 3 storeys

KO OK for soil type A



Lessons #1

Modern URM can be built in earthquake countries

Wrong analysis methods or Wrong codes = Wrong answers

Adequate models and commercial software, based on pushover analysis, are available for masonry structures with box behavior

Masonry Structures Without Box Behavior



Existing Masonry Structures “Existing built heritage often with high vulnerability: (a) very weak materials;

(b) heavy construction; (c) insufficient connections

Simple measures can help

A. Costa

FEUP+LNEC



Recent Tests: Flexible Diaphragm “Gaioleiro”-type structure (late 19th century / early 20th century)

Moderate damage for 100% of the design earthquake in Lisbon

Light strengthening and collapse for 150% of the design earthquake in Lisbon



Collapse at the base

Uniform Mass Distribution

Push-Over Analysis

Other Mass Distributions



Collapse: 4th balcony

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

0 5 10 15 20

Time [s]

Ac

ce

lera

tio

n [

m/s

2 ]

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0 0.5 1 1.5 2

Period [s]

Sei

smic

co

effi

cien

t A

h

DBERecord 1Record 2Record 3Record 4Record 5Série7

Second mode 0.50

10th mode 0.16

First mode 1.26

0

10

20

30

40

50

60

70

0 50 100 150 200 250 300

Maximum excentricity(Bending Moment / Axial Load )

H [

m]

Record 1Record 2Record 3Record 4Record 5Série6

FEM – Collapse for 0.20g

Five articificial accelerograms

REM

Time History Analysis



Modal Pushover

0

10

20

30

40

50

60

70

80

0 0.2 0.4 0.6

H [m

]

Maximum Displacement [m]

Rigid model

Beam model

Modal Pushover

0

10

20

30

40

50

60

70

80

0 2 4 6

H [m

]

Drift[%]

Rigid model

Beam model

Modal Pushover

90% of the total mass, 7 modes

Is adaptative pushover a solution?



Location: Lisbon

Material:Masonry walls andtimber pavements

No. of storeys: 4 to 6

Numerical model

“Gaioleiro” Building



“Gaioleiro” Building



Transversal Direction

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.01 0.02 0.03 0.04 0.05

Deslocamento [m]

Fac

tor

de

carg

a (λ

)

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.01 0.02 0.03 0.04 0.05

Displacement [m]

Sei

smic

coe

ffic

ien

t [α

h]

Pushover_1st ModePushover_MassDynamic

0

0.2

0.4

0.6

0.8

1

0 0.005 0.01 0.015 0.02

Deslocamento [m]

Fac

tor

de

carg

a [λ

]

0

0.2

0.4

0.6

0.8

1

0 0.005 0.01 0.015 0.02

Displacement [m]

Sei

smic

coe

ffic

ien

t [α

h]

Pushover_1st ModePushover_MassDynamic

Longitudinal Direction

Pushover Analysis



Numerical model

Principal strains

(external surface)

ε1

Experimental modelTime History Analysis



Adaptative Pushover Analysis

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.01 0.02 0.03 0.04 0.05

Deslocamento [m]

Fac

tor

de

carg

a (λ

)

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.01 0.02 0.03 0.04 0.05

Displacement [m]

Sei

smic

coe

ffic

ien

t [ɑ

h]PushoverDynamicSé i 3[ɑ

h]

Transversal Direction Longitudinal Direction



Design and Assessment = Macro-block analysis?

Limit equilibrium analysis using the principle of virtual work is currently understood as the “best” analysis technique

Overturning



Location: Guimarães

Style: Hybrid, with classical, gothic, renaissanceand romanic elements

Material: Granite stone masonry

Example



Cracking Pattern

Y Y’

Section YY’

Example



Cracking pattern

Main façade

Example



Cracking pattern

Views

Example



Geotechnical profiles

Example



FEM Model

Example



Soil structure-interaction

Example



Four possible mechanisms

α0 0.186

M* 4343.7 KN

e* 0.947 m/s2

Capacity a0* 0.197 g

Demand a0* 0.063 g

Safety Factor 3.13

Seismic Analysis



α0 0.184

M* 4254.5 KN

e* 0.953 m/s2


Demand a0* 0.086 g

Safety Factor 2.24

Seismic Analysis



α0 0.164

M* 8830.1 KN

e* 0.968 m/s2


Demand a0* 0.087 g

Safety Factor 1.94

Seismic Analysis



α0 0.205

M* 339.1 KN

e* 0.982 m/s2


Demand a0* 0.123 g

Safety Factor 1.69

Seismic Analysis



Método cinemático

Exemplo 2 – Santuário de SãoTorcato

Projecto de reforço



Rocking Motion and Multibody Dynamics

Masonry structures

Inverted pendulum(SDF problem)



Repeatability Test

Fourier SpectraHanning sine 3.3 Hz, 7 mm (P1)

0 1 2 3 4 5 6 7 8 9 10 11 12

-2

0

2

Roc

kin

g A

ng

le [

º]

Time [s]

Test 1 Test 2 Test 3 Test 4 Test 5 Test 6

0.1 1 10 1001E-4

1E-3

0.01

0.1

1

10

Am

plit

ud

e

Frequency [Hz]

Test 1 Test 2 Test 3 Test 4 Test 5 Test 6



Random Motion

Seismic record 20 – 1.1 (P2)

0 5 10 15 20 25-8

-6

-4

-2

0

2

4

6

8

10

12

Roc

king

Ang

le [

º]

Time [s]

Test 1 Test 2

0 5 10 15 20 25-8

-6

-4

-2

0

2

4

6

8

10

12

Roc

king

Ang

le [

º]

Time [s]

Test 1 Test 2

Seismic record 20 – 1.3 (P2)



Overturning Process

0.0 4 6 8 10 12 14-15

-10

-5

0

5

10

15

20

25

30

TransitionTransition

Roc

king

Ang

le [

º]

Time [s]

Critical AngleOverturning

Instability Return tostability

Specimen Critical Angle [º]

Theoretical Experimental Difference (%)

2 9.6 11.2 16

4 18.0 20.8 15

Full results, with:

• Experimental details• Numerical simulation• Stochastic analysis

On the dynamics of rigid-block structures: Applications to SDOFmasonry collapse mechanisms, PhD Thesis, Francisco Prieto



Multibody Dynamics

Stacked blocks and trilith Setup



Multibody Dynamics

Rocking patterns for stacked blocks

Final configurations for trilith



Lessons #2

Duration of the earthquake is a critical issue for brittle responses (be careful with scaling laws + do we know the duration of earthquakes?)

Repeatability does not hold for random analysis of brittle structures (probabilistic analysis is required)

Pushover methods fail to replicate observed failure modes of ancient masonry structures. Adaptive pushover and modal pushover analysis are not better. Macro-block can be a possible solution. But if you use them and fail to use the correct failure modes, the error can be large

If you use pushover analysis adopt a mass proportional load distribution (uniform horizontal acceleration)

Masonry Infills



A (very) old research issue

• Infill walls continue to fall. Life saving issue

• Estimated invested in masonry infills in Europe around 45 to 60 billioneuro. Greece (Parnitha, Magnitude 5.9, 1999), 60% of the repair costsdue to damage in masonry infills, associated finishings and installations(water, electricity, etc.). Insurance companies refers even higher costs(up to 80% of the total value of the building) for repairing andreconstructing non-structural elements. Cost issue

Facts



Reality

• L’Aquila earthquake (2009), Italy

Motivation



Scope of the project: Shaking table experimental program

Broader project and all the tests are based on the same geometry

Idealization of the geometry was done regarding the buildings constructed

in the last 20 years, in Portugal

Side View A Side View B

B

A5,7

33

3,23 3,23

Top View

6,45

Full scale geometry of study specimens (meters)



Preliminary Static Tests

Panel to be tested

In-plane (cyclic) tests Out-of-plane cyclic tests



Static tests (I)

After in-plane test

Out-of-plane movement (unreinforced vs. reinforced)

After out-of-plane test



In-plane tests Out-of-plane tests

-100

-75

-50

-25

0

25

50

75

-15 -10 -5 0 5 10 15

For

ça h

oriz

onta

l (kN

)

Deslocamento horizontal no plano (mm)-50

-25

0

25

50

-40 -20 0 20 40

For

ça h

orio

zont

al (

kN)

Deslocamento horizontal fora do plano (mm)

0

40

80

120

160

200

240

0 4 8 12

For

ça h

oriz

onta

l(k

N)

Deslocamento horizontal no plano (mm)

WALL_REF_01

WALL_JAR_02

Parede Simples

Parede Armada0

10

20

30

40

50

60

0 10 20 30 40 50 60

For

ça h

oriz

onta

l (kN

)

Deslocamento horiz. fora do plano (mm)

WALL_REF_01

WALL_RAR_02

Parede Simples

Parede Armada

Static tests (II)



Prototype reduced using Cauchy-Froude’s Simililitude Law to a scale 1:1,5

Three buildings were designed, each one with a different infill solution and

following either the Portuguese standards or Eurocodes (1,2 and 8)

One building represents what is built (unreinforced double leaf clay brick

masonry infill and older standards) and the other two future possibilities

(reinforced single leaf clay brick infill and Eurocodes)

Infill Solution 1 Infill Solution 2 Infill Solution 3




Shaking table test: Shape and dimensions of the model

East

South

West

North



Results: Video of stage 4



Results: Cracking pattern and collapse mode of infill walls

WestNorthStage 4

150% stage 3



Prototype reduced using Cauchy-Froud’s Similarity Law to a scale of 1:1,5












Prototype reduced using Cauchy-Froud’s Similarity Law to a scale of 1:1,5










Cracking in rendering around openings and corners

Results



Cracking in rendering around openings and corners

Results



Plaster removal: Toe and lintel crushing

Results



RC cracking (mid-height and joint)

Results

Separation between frame and infill



Conclusions

Rendering fell around the concrete frame, mainly at corners

Out-of-plane expulsion of infills prevented due to bed jointreinforcement or steel net connected to the concrete frame

The building, designed according to EC8 + reinforced infills,performed better than expected due to the masonry walls.No fragile collapses were found

The building designed according to the Portuguese old code+ unreinforced infills was able to perform better thanexpected due to the masonry walls. In the absence ofreinforcement (or other), out-of-plane expulsion was found,with a unacceptable collapse mode of the buildings

Few research exist on combined in-plane / out-of-plane.Demand is “unknown”. Capacity is almost unknown and theproblem is ill conditioned. We are looking for design rules



Lessons #3

Always connect the infills to the reinforced concrete frame (or find solutions to separate them and prevent out of plane failures)

Dynamic results can be unexpected. Be ready to change (First test was made without openings in the infill. Final tests were made with openings and additional masses). Cauchy-Froude’s similitude law required additional masses and a tedious addition of steel plates

Coming Up Tests



Masonry Infill Panels (I)



Masonry Infill Panels (II) Test masonry solutions from different parts of Europe

Test innovative masonry solutions (infill separation and dry-stack)

Test strengthened walls



Seismic performance of multi-storey timber buildings

Participants: University of Trento (Italy), University of Minho (Portugal), University ofGraz (Austria), Piú Legno (Italy), Rubner (Italy), Rusticasa (Portugal), Mayr-MelnhofKaufmann Gaishorn GmbH (Austria)

Construction system: Timber frame, log-house and cross-laminated timber







Larger Torsion Effects in Masonry (URM and Reinforced)



Dissipative Anchors

We are looking forward to more lessons…



Thank you for your attention

Recent Shaking Table Tests in Portugal: Lessons Learned 2012 earthquake engineering loure… ·...

Documents

Transcript of Recent Shaking Table Tests in Portugal: Lessons Learned 2012 earthquake engineering loure… ·...