Rel SERIES NTUA · PDF file‐ Final Report ‐ User Group ... countries with moderate and...
Transcript of Rel SERIES NTUA · PDF file‐ Final Report ‐ User Group ... countries with moderate and...
SEVENTH FRAMEWORK PROGRAMME
Capacities Specific Programme
Research Infrastructures
Project No.: 227887
SERIES SEISMIC ENGINEERING RESEARCH INFRASTRUCTURES FOR
EUROPEAN SYNERGIES
Work package [WP9 – TA5 LNEC]
Assessment of innovative solutions for non‐load bearing masonry enclosures
‐ Final Report ‐
User Group Leader: Elizabeth Vintzileou Revision: Final
July, 2013
SERIES 227887 MASONRY ENCLOSURES Project
i
ABSTRACT
This document reports the outcomes of the research project “Masonry Enclosures” developed in
the framework of the transnational access (TA) to LNEC shake table within the FP7 project
SERIES. This TA project addresses the seismic performance of masonry enclosures in European
countries with moderate and high seismicity and consisted on the experimental evaluation of the
seismic response of reinforced concrete (RC) frames with innovative solutions for masonry infill
walls, considering both the in-plane and out-of-plane behaviour of the enclosures.
In order to ensure that in-plane and out-of-plane damage of masonry infill walls due to seismic
actions complies with the performance levels’ requirements, Eurocode 8 imposes the use of
reinforced solutions. Nevertheless, it does not provide any design rules or methodologies for
such reinforced masonry enclosures. An experimental programme was thus defined for assessing
the response of innovative solutions for non-load bearing masonry enclosures using LNEC’s
triaxial shake table. This experimental programme comprised the testing of one RC frame
building and four independent wall panels. Both a horizontal reinforcement in the bedding planes
of the masonry units and a reinforced mortar coating solutions were tested on single leaf clay
brick infill walls. Furthermore, a testing device for masonry infill panels was specifically
conceived for this project. A detailed description of the methods used is given and the
experimental results are shown and interpreted on the basis of the structural response and its
evolution with damage.
Keywords: Non-load bearing masonry enclosures, reinforced concrete frames, bed joint reinforcement, wire mesh coating reinforcement, shaking table test, innovative test setup
SERIES 227887 MASONRY ENCLOSURES Project
ii
ACKNOWLEDGMENTS
The research leading to these results has received funding from the European Union Seventh
Framework Programme [FP7/2007-2013] under grant agreement n° 227887 [SERIES].
SERIES 227887 MASONRY ENCLOSURES Project
iii
REPORT CONTRIBUTORS
LNEC António Araújo Correia
Paulo Xavier Candeias
Alfredo Campos Costa
Ema Coelho
NATIONAL TECHNICAL UNIVERSITY Elizabeth Vintzileou
OF ATHENS Vasiliki Palieraki
UNIVERSITY OF MINHO Paulo B. Lourenço
João Leite
SERIES 227887 MASONRY ENCLOSURES Project
v
CONTENTS
1 Introduction .............................................................................................................................1
2 Description of the models and construction technique ............................................................4
2.1 Previous experimental studies ........................................................................................4
2.2 Previous tests at LNEC ...................................................................................................7
2.2.1 Results of model 1 ..............................................................................................7
2.2.2 Results of model 2 ............................................................................................25
2.3 Design and construction of the models .........................................................................37
2.3.1 Building model .................................................................................................37
2.3.2 Wall panels .......................................................................................................47
3 The LNEC Earthquake Engineering testing facility ..............................................................50
3.1 General information on the laboratory .........................................................................51
3.2 The facility: LNEC-3D Shaking Table .........................................................................51
3.3 General information on the shaking table .....................................................................53
3.4 Shaking table description ..............................................................................................53
3.5 Characteristics of the control system ............................................................................54
3.6 Complementary facilities ..............................................................................................55
4 Sensors technical data ............................................................................................................56
4.1 Displacement transducers .............................................................................................56
4.1.1 LVDT displacement transducers ......................................................................56
4.1.2 Hamamatsu optical system ...............................................................................57
4.1.3 Krypton K600 camera ......................................................................................58
4.2 Accelerometers .............................................................................................................60
4.3 Load cells ......................................................................................................................61
4.4 Acquisition system .......................................................................................................61
5 Test setup ...............................................................................................................................63
SERIES 227887 MASONRY ENCLOSURES Project
vi
5.1 Building model test setup .............................................................................................63
5.2 Wall panels test setup ...................................................................................................69
6 Seismic testing protocol ........................................................................................................76
6.1 Testing procedure .........................................................................................................76
6.2 Shaking table tuning .....................................................................................................76
6.3 Seismic test sequence ...................................................................................................79
7 Signal generation procedure for the shaking table tests ........................................................80
7.1 Building model .............................................................................................................80
7.2 Wall panels ...................................................................................................................83
8 Identification technique .........................................................................................................86
8.1 White noise ...................................................................................................................86
8.2 Impulse signal ...............................................................................................................87
9 Analysis of results .................................................................................................................89
9.1 Building model test results ...........................................................................................89
9.1.1 Initial test results ..............................................................................................89
9.1.2 Complementary test results ............................................................................100
9.1.3 Comparison of test results ..............................................................................109
10 Main conclusions .................................................................................................................119
References ....................................................................................................................................121
SERIES 227887 MASONRY ENCLOSURES Project
vii
List of Figures
Figure 1.1: Reduced scale RC building .......................................................................................... 2
Figure 1.2: Details of reinforced mortar coating ............................................................................ 3
Figure 1.3: Geometry of the model ................................................................................................. 3
Figure 1.4: Wall panels test setup ................................................................................................... 3
Figure 2.1: Test setup in [1]: (a) in-plane test; (b) out-of-plane test ............................................... 4
Figure 2.2: Model detailing in [2] and [4]: (a) infill wall solutions; (b) detail of the mortar
coating reinforcement ..................................................................................................................... 5
Figure 2.3: Testing setup and models [17]: (a) in-plane test; (b) out-of-plane test; (c) model 1; (d)
model 2; (e) model 3 ....................................................................................................................... 6
Figure 2.4: Position and label of the accelerometers in model 1 .................................................... 7
Figure 2.5: Crack patterns of the exterior leaf of model 1 after stage 3 (2475 YRP) (Note: the
drawn lines on the RC frame represent damage on the clay bricks applied to the RC frame to
avoid thermal bridges) .................................................................................................................... 9
Figure 2.6: Crack patterns of the interior leaf of model 1 after stage 3 (2475 YRP) (Note: the
drawn lines on the RC frame represent damage on the clay cricks applied to the RC frame to
avoid thermal bridges) .................................................................................................................... 9
Figure 2.7: Stage 4 (4574 YRP) of the shaking table test of model 1: (a) out-of-plane collapse of
the exterior leaf of the infill wall at the ground floor of the South façade; (b) out-of-plane
collapse of the interior leaf of the infill wall at the ground floor of the South façade; (c) out-of-
plane collapse of the exterior jambs of the infill walls at the first storey of the East façade; (d)
out-of-plane collapse of both leaves of the infill wall at the ground storey of the North façade; (e)
model 1 after the fourth stage, collapsed and without all the infill walls of the ground floor; (f)
ground floor column of the Northwest after collapsing at the top and disintegration up to mid-
height; (g) plastic hinge developed on the top of the ground RC column of the Northeast corner;
(h) barely damaged infill wall at the first storey of the South façade. .......................................... 11
Figure 2.8: Plastic hinge formation at the top of the ground floor columns ................................. 12
Figure 2.9: Mode shapes of the DI 0 of model 1 (initial dynamic identification test) .................. 15
SERIES 227887 MASONRY ENCLOSURES Project
viii
Figure 2.10: Frequency change along loading stages: (a) variation of the FRF’s along the test of
model 1 at the accelerometer BNE – 1T; (b) evolution of the frequencies along the test of model
1 and their final variation in respect to DI 0 ................................................................................. 15
Figure 2.11: Seismic vulnerability curves of model 1 in the transverse and longitudinal
directions, using the PGA and the Input Energy as input. Here, the damage indicator is a measure
of the frequency change ................................................................................................................ 17
Figure 2.12: Evolution of the frequencies of the infill walls in the North and South façades along
the test of model 1 and their final variation in respect to DI 0. .................................................... 19
Figure 2.13: Interstorey displacements and drifts of model 1 ...................................................... 20
Figure 2.14: Recorded PGA and amplifications at the infill walls and at the RC structure for each
test stage of model 1 ..................................................................................................................... 22
Figure 2.15: Out-of-plane deformation of the North and South infill walls along the tests of
model 1.......................................................................................................................................... 24
Figure 2.16: Out-of-plane PGD of the East and West infill walls of model 1 .............................. 25
Figure 2.17: Position and label of the accelerometers in model 2 ................................................ 25
Figure 2.18: Crack patterns of model 2 after stage 3 (2475 YRP) (Note: the drawn lines on the
RC frame represent damage on the rendering applied to the RC frame) ...................................... 27
Figure 2.19: Crack patterns of model 2 after stage 4 (4574 YRP) (Notes: the drawn lines on the
RC frame represent damage on the rendering applied to the RC frame. The blue lines developed
after stage 3) .................................................................................................................................. 27
Figure 2.20: Damage in model 2 after the fourth stage (4574 YRP): (a) North façade; (b) South
façade; (c) West façade from the inside; (d) detail of the left jamb of the door on the North
façade and a horizontal crack at mid-height of the Northeast corner column; (e) horizontal crack
at mid-height of the Southwest corner column; (f) heavily damaged top column-beam connection
of the Southwest corner column with loss of the concrete cover and rebar exposure .................. 28
Figure 2.21: Mode shapes of the DI 0 of model 2 (initial dynamic identification test) ................ 29
Figure 2.22: Frequency change along loading stages: (a) variation of the FRF’s along the test of
model 2 at the accelerometer BNE – 2T; (b) evolution of the frequencies along the test of model
2 and their final variation in respect to DI 0 ................................................................................. 30
Figure 2.23: Seismic vulnerability curves of model 2 in the transverse and longitudinal
directions, using the PGA and Input Energy as input ................................................................... 31
SERIES 227887 MASONRY ENCLOSURES Project
ix
Figure 2.24: Evolution of the frequencies of the infill walls in the North and South façades along
the test of model 2 and their final variation in respect to DI 0 ..................................................... 32
Figure 2.25: Interstorey displacements and drifts of model 2 ...................................................... 33
Figure 2.26: Recorded PGA and amplifications at the infill walls and at the RC structure for each
test stage of model 2 ..................................................................................................................... 34
Figure 2.27: Out-of-plane deformation of the infill wall at the ground level of the South façade
(mm) .............................................................................................................................................. 35
Figure 2.28: Out-of-plane deformation of the infill wall at the first storey of the South façade
(mm) .............................................................................................................................................. 36
Figure 2.29: Out-of-plane deformation of the infill wall at the first storey of the North façade
(mm) .............................................................................................................................................. 36
Figure 2.30: Out-of-plane PGD of the North, East and West infill walls (mm) ........................... 37
Figure 2.31: Prototype geometry (m)............................................................................................ 37
Figure 2.32: Geometry of the tested model reduced to a scale of 1:1.5 ....................................... 39
Figure 2.33: Response spectra after the application of the similitude law, see Table 2.2,
according to EC8 [29] ................................................................................................................... 41
Figure 2.34: Geometry of the openings in each façade: (a) North façade; (b) West façade; (c)
East façade; (d) South façade ........................................................................................................ 42
Figure 2.35: Additional steel masses for: (a) RC concrete structure, bolted to the slabs of the 1st
floor and roof with 82x82x26 cm and 12KN each; (b) Infill walls, bolted to both sides of the wall
with 15x15x4 cm and 0.072KN each ............................................................................................ 43
Figure 2.36: Construction of the models: (a) horizontally aligned surface on which the models
were constructed; (b) RC ring beam with steel connector with an eye in lift and transport the
model to the shaking table; (c) model 1 on the shaking table before the test ............................... 44
Figure 2.37: Single leaf clay brick infill walls with reinforced plaster from Model 3 already
scaled: (a) spacing of the Hilti X-M8H10-37-P8 connectors along the height of the RC column;
(b) detail at the RC column; (c) detail of the Hilti X-M8H10-37-P8 connectors ......................... 45
Figure 2.38: Construction of the infills of model 3: (a) Bekaert Armanet ϕ1.05mm 12.7x12.7mm;
(b) Hilti X-M8H10-37-P8; (c) application of the grid in the outer surface at a corner column; (d)
application of the grid in the inner surface; (e) additional masses with steel rings attached to the
infill walls ..................................................................................................................................... 46
Figure 2.39: Frame panel of typical RC building ......................................................................... 47
SERIES 227887 MASONRY ENCLOSURES Project
x
Figure 2.40: Reinforcement layout for the RC frames ................................................................. 47
Figure 2.41: Prestressing details ................................................................................................... 48
Figure 2.42: RC frame ready for infill construction ..................................................................... 48
Figure 2.43: Masonry infill construction with bed joint reinforcement ........................................ 48
Figure 2.44: Specimens ready for testing ..................................................................................... 49
Figure 3.1: LNEC Earthquake Engineering testing facility (Ferry Borges building) ................... 50
Figure 3.2: LNEC earthquake engineering testing room .............................................................. 52
Figure 3.3: LNEC-3D shaking table ............................................................................................. 52
Figure 4.1: LVDT displacement transducers (source: RDP) ........................................................ 56
Figure 4.2: HAMAMATSU optical 2D displacement transducer ................................................ 57
Figure 4.3: Hamamatsu system configuration [12] ....................................................................... 58
Figure 4.4: The Krypton K600 camera [19] ................................................................................. 58
Figure 4.5: Representative measurement volume of a Krypton K600 camera [19] ..................... 59
Figure 4.6: Accuracy zones of the Krypton K600 camera system [19] ........................................ 59
Figure 4.7: Endevco accelerometers ............................................................................................. 60
Figure 4.8: PCB Piezotronics accelerometers ............................................................................... 60
Figure 4.9: Instron load cells ........................................................................................................ 61
Figure 5.1: Accelerometers used in the shaking table tests: (a) piezoelectric from PCB [31], [32];
(b) piezoelectric from Wilcoxon [40]; (c) capacitive from Endevco [9] ...................................... 64
Figure 5.2: Accelerometers setup: (a) North and East façades; (b) South and West façades ....... 65
Figure 5.3: PCB Piezotronics accelerometers: (a) at the infill walls; (b) at the corners of the RC
slabs............................................................................................................................................... 65
Figure 5.4: Hamamatsu photonics c5949 [12]: (a) position of the Hamamatsu leds in the first
storey; (b) position of the Hamamatsu leds in the roof; (c) camera and led at the corner of the
structure; (d) infrared led; (e) controller ....................................................................................... 66
Figure 5.5: Acquisition and control room: (a) from top to bottom: NI-SCXI-1001, PCB
Piezotronics 481A02 and NI PXI-1052; (b) control room with the shaking table’s controls and
the model’s acquisition system ..................................................................................................... 67
Figure 5.6: K600 Krypton camera: (a) three CCD cameras; (b) Space Probe used to calibrate the
initial geometrical plan of the LED’s; (c) acquisition control; (d) distribution of the LED’s along
the infill wall on the upper floor of the North façade ................................................................... 68
Figure 5.7: Main steel caisson frames of TIM (construction phase) ............................................ 69
SERIES 227887 MASONRY ENCLOSURES Project
xi
Figure 5.8: Base columns of main steel frame with detail of bolted connection to the shake table
(construction phase) ...................................................................................................................... 69
Figure 5.9: Guiding system of RC frame upper beam and rollers for longitudinal motion
(construction phase) ...................................................................................................................... 70
Figure 5.10: Pyramidal support for strut connection between the RC frame and the reaction wall
....................................................................................................................................................... 71
Figure 5.11: Hinged base supports for the RC frame specimens (construction phase) ................ 71
Figure 5.12: Hinged base and pyramidal supports on their final position .................................... 71
Figure 5.13: Assembly of TIM components on the shaking table ................................................ 72
Figure 5.14: Positioning of TIM over the wall panel setup .......................................................... 72
Figure 5.15: Complete setup for wall panels tests ........................................................................ 72
Figure 5.16: Schematic representation of the finite element models used for the design of TIM,
taking into account (right) or not (left) the wall panels ................................................................ 73
Figure 5.17: Vibration modes of TIM without the wall panel contribution: a) longitudinal
(f = 19.9 Hz); b) transverse (f = 33.8 Hz) ..................................................................................... 73
Figure 5.18: Vibration modes of TIM with the wall panel contribution: a) longitudinal
(f = 18.4 Hz); b) transverse (f = 23.1 Hz); c) torsional (f = 25.5 Hz) ........................................... 73
Figure 5.19: Out-of-plane wall panel deformation monitoring with Krypton K600 camera ........ 74
Figure 5.20: Video camera and target points for data image correlation measurement of in-plane
deformations at one RC frame node ............................................................................................. 74
Figure 5.21: Hamamatsu setup for measuring the horizontal translations of the RC frame nodes
....................................................................................................................................................... 75
Figure 5.22: Accelerometer setup for RC frame out-of-plane vibration measurements ............... 75
Figure 5.23: Load cells for strut reaction measurement ............................................................... 75
Figure 6.1: Shaking table tuning application: definition of parameters ........................................ 77
Figure 6.2: Shaking table tuning application: FRF obtained ........................................................ 77
Figure 6.3: Signal tuning iterative process ................................................................................... 78
Figure 6.4: Calibration of the input signals with masses attached to the shaking table ............... 78
Figure 7.1: Comparison between pseudo-acceleration response spectra of the accelerograms
generated and the response spectra, already scaled following the similitude law of Cauchy-
Froude, obtained from EC8 [30] ................................................................................................... 81
SERIES 227887 MASONRY ENCLOSURES Project
xii
Figure 7.2: Time histories of the input signal of stage 2 (SD 475 YRP) reduced at 1:1.5 scale
using Cauchy-Froude’s similitude law (see Table 2.2) ................................................................ 82
Figure 7.3: Comparison between pseudo-acceleration response spectra of the accelerograms
generated and the response spectra obtained from EC8 [30] ........................................................ 83
Figure 7.4: Representative building model for wall panel input time-history definition ............. 83
Figure 7.5: Longitudinal modes of vibration 1 (1.35Hz) and 2 (4.28Hz) ..................................... 84
Figure 7.6: Transverse modes of vibration 1 (2.78Hz) and 2 (10.94Hz) ...................................... 84
Figure 7.7: Interstorey drift time-history for in-plane motion ...................................................... 85
Figure 7.8: Absolute acceleration time-history for out-of-plane motion ...................................... 85
Figure 8.1: White noise signals for dynamic identification tests .................................................. 87
Figure 9.1: Position and label of the accelerometers in model 3 .................................................. 89
Figure 9.2: Crack patterns of model 3 after stage 2 (475 YRP) (Note: the drawn lines on the RC
frame represent damage on the mortar rendering applied to the RC frame) ................................ 90
Figure 9.3: Crack patterns of model 3 after stage 3 (2475 YRP) (Note: the drawn lines on the RC
frame represent damage on the mortar rendering applied to the RC frame) ................................ 91
Figure 9.4: Damage in model 3 after stage 3 (2475 YRP): (a) infill wall at the ground floor of the
North façade; (b) infill wall at the upper floor in the East façade; (c) damaged mortar rendering
at the Southeast corner; (d) damaged mortar rendering at the Southwest corner ......................... 91
Figure 9.5: Mode shapes of the DI 0 of model 3 (initial dynamic identification test) .................. 93
Figure 9.6: Frequency change along loading stages: (a) variation of the FRF’s along the test of
model 3 at the accelerometer BNE – 1L; (b) evolution of the frequencies along the test of model
3 and their final variation in respect to DI 0 ................................................................................. 93
Figure 9.7: Seismic vulnerability curves of model 3 in the transverse and longitudinal directions,
using the PGA and Input Energy as input ..................................................................................... 94
Figure 9.8: Evolution of the frequencies of the infill walls in the North and South façades along
the test of model 3 and their final variation in respect to DI 0 ..................................................... 95
Figure 9.9: Interstorey displacements and drifts of model 3 ........................................................ 96
Figure 9.10: Recorded PGA and amplifications at the infill walls and at the RC structure for each
test stage of model 3 ..................................................................................................................... 97
Figure 9.11: Out-of-plane deformation of the infill wall at the ground level of the South façade of
model 3 in mm .............................................................................................................................. 99
SERIES 227887 MASONRY ENCLOSURES Project
xiii
Figure 9.12: Out-of-plane deformation of the infill wall at the first storey of the South façade of
model 3 in mm .............................................................................................................................. 99
Figure 9.13: Out-of-plane deformation of the infill wall at the first storey of the North façade of
model 3 in mm. ........................................................................................................................... 100
Figure 9.14: Out-of-plane PGD of the North, East and West infill walls of model 3 in mm ..... 100
Figure 9.15: Damage in model 3B after stage 3 (2475 YRP): (a) North façade; (b) South façade;
(c) crack and mortar rendering loss at the Northwest corner; (d) crack and mortar rendering loss
at Northeast corner; (e) crack at the a lateral jamb in the infill wall at the ground floor of the East
façade; (f) crack at the interior jambs in the infill wall at the ground floor of the East façade .. 102
Figure 9.16: Damage in the infill walls and RC structure after the reinforced rendering removal
at the ground floor: (a) infill wall of the North façade; (b) South infill wall with a compression
crush at right down corner; (c) gap between one of the West the infill wall and RC frame in the
West wall; (d) infill walls of the West façade; (e) extensive cracking at the upper column-beam
connection in the Northwest corner ............................................................................................ 103
Figure 9.17: Evolution of the frequencies along the test of model 3B, and their final variation in
respect to DI 0 of model 3, at the RC structure and infill walls in South façade and ground level
of the North façade...................................................................................................................... 104
Figure 9.18: Interstorey displacements and drifts of model 3B .................................................. 105
Figure 9.19: Recorded PGA and amplifications at the infill walls and at the RC structure for each
test stage of model 3B ................................................................................................................. 106
Figure 9.20: Out-of-plane deformation of the infill wall at the ground level of the South façade of
model 3B in mm.......................................................................................................................... 107
Figure 9.21: Out-of-plane deformation of the infill wall at the first storey of the South façade of
model 3B in mm.......................................................................................................................... 108
Figure 9.22: Out-of-plane deformation of the infill wall at the first storey of the North façade of
model 3B in mm.......................................................................................................................... 108
Figure 9.23: Out-of-plane PGD of the North, East and West infill walls of model 3B in mm .. 109
Figure 9.24: Longitudinal direction target/acquired comparison (Fourier filter: 1-40Hz): (a)
PGA; (b) Root Mean Square Acceleration; (c) Arias Intensity: (d) Input Energy ..................... 112
Figure 9.25: Transverse direction target/acquired comparison (Fourier filter: 1-40Hz): (a) PGA;
(b) Root Mean Square Acceleration; (c) Arias Intensity: (d) Input Energy ............................... 112
SERIES 227887 MASONRY ENCLOSURES Project
xiv
Figure 9.26: Vulnerability curves of the 1st mode in each main direction of the RC structure of
the three tested models ................................................................................................................ 113
Figure 9.27: Average vulnerability curves of the North and South infill walls of the three tested
models ......................................................................................................................................... 114
Figure 9.28: Interstorey displacements of the three tested models in the transverse and
longitudinal directions ............................................................................................................... 116
Figure 9.29: Interstorey drifts of the three tested models in the transverse and longitudinal
directions ..................................................................................................................................... 116
Figure 9.30: Average recorded PGA and amplifications at the infill walls and at the RC structure
for each test stage of all tested models ........................................................................................ 118
SERIES 227887 MASONRY ENCLOSURES Project
xv
List of Tables
Table 2.1 - Experimental damping ratio of model 1 ..................................................................... 16
Table 2.2 - Cauchy-Froude similitude law ................................................................................... 39
Table 2.3 - Design loads of the models already reduced at scale of 1:1.5 .................................... 40
Table 3.1 – Name and location of the Laboratory ........................................................................ 51
Table 3.2 – Name and location of the parent organization ........................................................... 51
Table 3.3 – Name of the LNEC-3D shaking table ........................................................................ 53
Table 3.4 – Type of shaking table ................................................................................................. 53
Table 3.5 – Characteristics of the Platform .................................................................................. 53
Table 3.6 – Characteristics of the Actuators ................................................................................. 53
Table 3.7 – Shaking table performances ....................................................................................... 54
Table 3.8 – Characteristics of the analogue part ........................................................................... 54
Table 3.9 – Characteristics of the digital part ............................................................................... 54
Table 4.1 – Characteristics of the RDP displacement transducers ............................................... 57
Table 4.2 – Characteristics of the HAMAMATSU displacement transducers ............................. 57
Table 4.3 – Characteristics of the Endevco accelerometers ......................................................... 60
Table 4.4 – Characteristics of the PCB accelerometers ................................................................ 60
Table 4.5 – NI PXI controller ....................................................................................................... 61
Table 4.6 – NI PXI chassis ........................................................................................................... 62
Table 6.1 - Shaking table test procedure for the building model .................................................. 79
Table 9.1 - Experimental damping ratios of model 3 ................................................................... 94
SERIES 227887 MASONRY ENCLOSURES Project
1
1 Introduction
The masonry enclosures project focused on the seismic performance assessment of traditional
and innovative masonry enclosures of European countries with moderate to high seismicity
(Greece, Portugal, Italy and Slovenia). A complete redesign of the experimental program was
undertaken in order to potentiate the goals of the project which were to characterize the
seismic behaviour of different types of traditional and immediate future systems for masonry
enclosures.
In fact, the recent L’Aquila earthquake of 2009 has reminded seismic engineers that the
current masonry infill solutions are not effective, as illustrated by the considerable in-plane
damage and out-of-plane collapses verified in a common basis throughout the affected areas.
Eurocode 8 addresses this issue by imposing the use of reinforced infill solutions but fails to
give design and detailing methodologies.
With the above in mind, a shake table experimental research programme was devised in order
to investigate the seismic behaviour of reinforced infill walls and how they affect the global
structural response.
Four types of masonry enclosures were intended to be tested within this TA project:
i) Unreinforced masonry
ii) Horizontal layers of reinforcement between masonry units
iii) Reinforced mortar coating
iv) Both a reinforced mortar coating and horizontal layers of reinforcement between
masonry units
The first phase of the research activity involved the seismic testing of a two-storey reinforced
concrete (RC) infilled frame building designed to the Eurocodes and built at a 1:1.5 scale. As
a follow up of previous tests performed at LNEC using RC buildings either with double leaf
unreinforced masonry infills or with single leaf masonry with bed joint reinforcement, this
reduced scale model was built with single leaf clay bricks and reinforced mortar coating, as
shown in Figure 1.1 to Figure 1.3. Wire mesh reinforcement was placed on both sides of the
SERIES 227887 MASONRY ENCLOSURES Project
2
infill walls and anchored to the RC frame and masonry units. Additional masses were
attached to the walls in order to respect Cauchy-Froude similitude requirements. From these
tests it was possible to assess the evolution of the seismic behaviour of infills and their
influence on the RC structure through several acceleration inputs of increasing amplitude,
associated to cumulative damage limit states.
The second phase of this transnational access activity comprised the dynamic testing of four
specimens of a closed RC plane frame with external dimensions of 6.40mx3.25m. This plane
frame was tested simultaneously for in-plane and out-of-plane dynamic actions, representing
the response of a typical upper storey frame panel of a RC building. Both motions match
realistic floor response spectra, of narrow band frequency content. The in-plane motion
enforces an inter-storey drift time-history on the frame by restraining the upper beam and by
imposing the displacement of the shaking table on the lower beam. On the other hand, the out-
of-plane motion consists on a rigid-body vibration of both the upper and lower beams,
reproducing the storey absolute accelerations and thus inducing high-frequency inertia forces
perpendicular to the masonry panel and leading to a local vibration of the infill wall.
This unique testing setup (Figure 1.4) was specifically designed for this test and is mainly
composed of a stiff steel caisson three-dimensional frame which moves rigidly with the
shaking table. It is fixed to the upper beam in the transversal direction, while a system of
rollers allows for an independent motion in the longitudinal direction.
Figure 1.1: Reduced scale RC building
SERIES 227887 MASONRY ENCLOSURES Project
3
Figure 1.2: Details of reinforced mortar coating
Figure 1.3: Geometry of the model
Figure 1.4: Wall panels test setup
SERIES 227887 MASONRY ENCLOSURES Project
4
2 Description of the models and construction
technique
2.1 PREVIOUS EXPERIMENTAL STUDIES
The interaction between the infill wall and the surrounding frame has been studied by several
authors since the 1960’s (see [13]; [15]; [16]; [23]; [24]; [25]; [39]; [40]). Given the
objectives of the present work, more recent studies that relate the in-plane and out-of-plane
damage, or that use dynamic actions are of higher relevance. In [1], the out-of-plane
behaviour of RC frames with infill walls, after damaging the frame in-plane by applying a
horizontal load, was evaluated. Eight RC frames were tested during the experimental
program, at a 1:1 scale, using the test setup in Figure 2.1, and the main conclusions were: (i)
the in-plane stiffness of the system is directly proportional to the compressive strength of the
masonry, and it drops significantly after the first crack; (ii) the out-of-plane capacity depends
on the slenderness of the wall and on the compressive strength of the masonry; (iii) for large
slenderness, the out-of-plane capacity of the infill wall decreases after being damaged in-
plane.
(a) (b)
Figure 2.1: Test setup in [1]: (a) in-plane test; (b) out-of-plane test
SERIES 227887 MASONRY ENCLOSURES Project
5
In the work described in [10], the authors tested masonry infill walls confined in metallic
frames for in-plane, out-of-plane and combined actions with the objective of understanding
the seismic behaviour of damaged infills. The authors also intended to identify the least
favourable combination of actions, for which reason the combined actions test was performed
in different sequences. The in-plane test was carried out by imposing cyclic incremental
displacements. The main conclusions were: (i) damaged infill walls have a large out-of-plane
capacity in spite of presenting larger deformations when compared to undamaged ones; (ii)
the out-of-plane damage in the infill walls does not affect the maximum in-plane compressive
strut capacity.
In [2] and [4], the authors tested RC frames with both reinforced and unreinforced infill walls
loaded initially in-plane and then out-of-plane, with the objective of assessing the seismic
capacity of different reinforcement solutions. The frames were built at 1:1 scale, with in-plane
dimensions of 4.2mx3.0m, according to the solutions described in Figure 2.2.
(a) (b)
Figure 2.2: Model detailing in [2] and [4]: (a) infill wall solutions; (b) detail of the mortar coating reinforcement
The test plan for each frame consisted in the application of a vertical load to the columns, kept
constant during the test in order to simulate the load transferred by the upper storeys, followed
by a cyclic in-plane drift, from 0.1% to 3.6%. The infill wall was then loaded out-of-plane
with a monotonic load applied in four different points. The test results led to the following
conclusions: (i) the reinforcement reduces the in-plane damage but it does not considerably
increase the stiffness, when compared to the unreinforced solution; (ii) damaged unreinforced
infill walls needed an acceleration five times lower to collapse out-of-plane, when compared
SERIES 227887 MASONRY ENCLOSURES Project
6
to the undamaged infill wall; (iii) the presence of reinforcement increases the out-of-plane
stiffness of the infill wall, hence smaller displacements were recorded.
In the experimental campaign presented in [17], the author tested RC frames with infill
masonry walls, at a 1:2 scale, by applying in-plane cyclic horizontal loads to the frame
followed by out-of-plane accelerations imposed by a uniaxial shaking table, as represented in
Figure 2.3. The objective was to understand the combined seismic behaviour of a simple and
slender solution (model 1), a solution with an RC lintel and column at mid-span (model 2)
and a solution with a more robust RC frame (model 3). The conclusions of the work were: (i)
the more slender models presented the highest inertial force at mid-height while the more
robust model presented the highest value at the top; (ii) the additional RC members in model
2 improved its seismic behaviour by reducing the out-of-plane displacements and through a
slower crack spreading process; (iii) the out-of-plane collapse is dependent not only on the
corresponding inertial forces but also on the excessive out-of-plane displacements.
(a) (b)
(c) (d) (e)
Figure 2.3: Testing setup and models [17]: (a) in-plane test; (b) out-of-plane test; (c) model 1; (d) model 2; (e) model 3
SERIES 227887 MASONRY ENCLOSURES Project
7
2.2 PREVIOUS TESTS AT LNEC
Two building models, similar to the one included in this project, were previously tested at the
Earthquake Engineering and Structural Dynamics Division (NESDE) of the National
Laboratory for Civil Engineering (LNEC). Their description and test results are described in
the following sections.
2.2.1 Results of model 1
Model 1 was designed following the pre-Eurocode normative, RSA [36] and REBAP [35],
using the most commonly used concrete and steel for rebars (C20/25 and S400, respectively),
together with double leaf, unreinforced, clay brick infill walls. Hence, model 1 represents the
built heritage in Portugal for the last three decades.
The following results were obtained using the acquisition equipment described in Chapter 4:
data recordings of the tests (quantitative results) and damage maps drawn between each of the
test stages (qualitative results). Figure 2.4 presents the position and label of the
accelerometers in model 1. Since the clay brick infills have two leaves, a set of accelerometers
was placed in the interior leaf at the exact same position of the exterior accelerometers seen in
Figure 2.4, for comparison purposes. The label of the interior accelerometers was obtained by
replacing the E with an I.
North South East West
Figure 2.4: Position and label of the accelerometers in model 1
NE1 1
BNE 2L
NE2 2
NE1 2
INP L
NE1 3
NE2 3
BNE 1L
NE2 1
SE1 3
SE2 2
BSW 1L
SE1 1
SE2 1 SE2 3
SE1 2
BSW 2L
EE1.21
EE1.11
BNE 1T
EE2.2 2
BNE 2T
EE2.1 1
EE2.2 1
WE1.1 1
BSW 2T
BSW 1T
WE1.2 1
INP T
SERIES 227887 MASONRY ENCLOSURES Project
8
Overall damage and crack patterns
The test procedure followed in the present work, which is the input of four seismic actions of
increasing amplitude, leads to damage accumulation. The evolution of the damage can be a
strong indicator of the collapse mechanism developed by the structure, especially in model 1
due to its collapse during the last stage of the shaking table test. Furthermore, the analysis and
relation of the crack patterns with the quantitative results, obtained from the data acquisition
equipment as accelerometers and displacement measurement cameras, will increase the
reliability of conclusions.
Even though the model was transported to the shaking table using a crane, the model did not
present any noticeable damage before the first stage of the shaking table test. After the first
two stages (225 and 475 YRP), model 1 did not present any visible damage, which is not in
agreement with the dynamic data that shows a small decrease in the model frequencies. This
loss of stiffness can be related to two aspects: i) the separation of the infill walls from the
reinforced concrete (RC) frame, a damage that is difficult to detect and is camouflaged by the
mortar rendering of the infill walls; ii) cracks and micro cracks in the RC frame that remain
undetected due to the clay bricks applied externally to avoid thermal bridges. As expected,
after the third stage (2475 YRP), the model presented clear cracks on both leaves of the infill
walls, see Figure 2.5 and Figure 2.6, mainly at the ground storey of the North, East and West
façades. The infill wall at the first storey of the North façade also presented some cracks. The
cracks appeared mainly at the connection between the infill wall and the RC frame, and at the
corners of the openings and moving towards the RC frame. In the infill walls at the ground
floor in the East and West façade, and on both leaves, the crack pattern around several
opening jambs is clear, separating them from the RC frame and the section of the infill wall
below the openings. This damage is related to the in-plane displacements of the RC frames in
the North-South (longitudinal) direction. Associated with this damage, the frequencies of the
first three mode shapes decreased 13.6%, 28.4% and 20.2%, respectively, in comparison to
the undamaged state.
SERIES 227887 MASONRY ENCLOSURES Project
9
North South
East West
Figure 2.5: Crack patterns of the exterior leaf of model 1 after stage 3 (2475 YRP) (Note: the drawn lines on the RC frame represent damage on the clay bricks applied to the RC frame to avoid thermal bridges)
North South
East West
Figure 2.6: Crack patterns of the interior leaf of model 1 after stage 3 (2475 YRP) (Note: the drawn lines on the RC frame represent damage on the clay cricks applied to the RC frame to avoid thermal bridges)
Model 1 collapsed during the fourth and last stage (4574 YRP), after losing the infill walls,
see Figure 2.7 (a) to (d), with subsequent failure of the three RC columns at the ground storey
of the West façade, see Figure 2.7 (e) and (f). The collapse mechanism developed, designated
SERIES 227887 MASONRY ENCLOSURES Project
10
here by soft storey, is characterized by the concentration of plastic hinges at the columns of a
given storey, while the upper part remains rather stiff, Figure 2.7 (g), is very undesirable
during a seismic action since it commonly leads to the partial or total collapse of the structure,
as it was the case in the present test. A beam sway mechanism, where the plastic hinges are
developed at the beams and not at the columns, is more desirable since it dissipates the energy
transferred by the earthquake without compromising its stability [40]. The collapse of the
columns occurred at their top, in the RC node, see Figure 2.7 (g) and Figure 2.8, followed by
disintegration of the concrete and instability of the steel up to mid-height of the column. This
failure further stresses the need to adequately confine concrete in the nodes and the need to
add more stirrups to avoid shear failure. It seems that the concentration of damage and
deformation of the columns in the nodes is also forced by the stiff behaviour of the first storey
and, possibly, the presence of the masonry infills in the ground storey, before collapse.
Before the collapse of the structure, the central jambs at the first storey of the East façade
collapsed out-of-plane, see Figure 2.7 (c), followed by the infill wall at the ground storey of
the North façade, see Figure 2.7 (d). The exterior leaf of the infill wall at the ground floor of
the South façade and the infill walls at the ground floor of the East and West façade collapsed
out-of-plane simultaneously, see Figure 2.7 (a).
SERIES 227887 MASONRY ENCLOSURES Project
11
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Figure 2.7: Stage 4 (4574 YRP) of the shaking table test of model 1: (a) out-of-plane collapse of the exterior leaf of the infill wall at the ground floor of the South façade; (b) out-of-plane collapse of the interior leaf of the infill wall at the ground floor of the South façade; (c) out-of-plane collapse of the
exterior jambs of the infill walls at the first storey of the East façade; (d) out-of-plane collapse of both leaves of the infill wall at the ground storey of the North façade; (e) model 1 after the fourth stage,
collapsed and without all the infill walls of the ground floor; (f) ground floor column of the Northwest after collapsing at the top and disintegration up to mid-height; (g) plastic hinge developed on the top of the ground RC column of the Northeast corner; (h) barely damaged infill wall at the first storey of the
South façade.
All these infills collapsed with a rotation mechanism with a hinge line at their bottom support
or at the first masonry joint (as a cantilever). The interior leaf of the infill wall at the ground
floor of the South façade was the last infill to collapse, see Figure 2.7 (b). This infill collapsed
SERIES 227887 MASONRY ENCLOSURES Project
12
with three hinge lines (top, centre and bottom). Immediately after, the structure collapsed. The
jambs around the windows collapsed usually by rotating out-of-plane as a rigid body with a
hinge line close to the connection to the spandrel (either the support or the first masonry
joint), or the rest of the masonry, again as a cantilever.
Figure 2.8: Plastic hinge formation at the top of the ground floor columns
SERIES 227887 MASONRY ENCLOSURES Project
13
Modal frequencies of the RC structure
Model 1 was subjected to four dynamic identification tests, from DI0 (undamaged state) to
DI3 (after stage 3). The model collapsed during stage 4, therefore it was not possible to
perform the last dynamic identification test. The quality of the obtained results can be
measured by the coherence between the input and output signals, which should be close to 1,
and this was the case in all the DI tests. Clear peaks could also be identified in the FRF’s
(Frequency Response Functions).
Five mode shapes were identified in DI0, see Figure 2.9, namely: the first and second
transverse modes; the first and second longitudinal modes; the (first) torsional mode. As
expected, the first mode is transverse (East-West) at a frequency of 7.71Hz, as the RC frames
in that direction are single-bay and the total length of the model is smaller than in the
longitudinal direction. The second mode is longitudinal (North-South) at the frequency of
9.62Hz since the RC frames are double bay and the total length of the model is higher than
the transverse one. Due to influence of the infill walls, and the fact that the percentage of
openings is not the same in all façades, the first transverse and longitudinal modes have a very
small component in the longitudinal and transvers directions, respectively. As it can be seen
in Figure 2.10 (a), the first mode shape was clearly identified in the FRF.
The torsional mode has a frequency of 26.95Hz, considerably higher than the previous
identified modes. This increment in the global torsional stiffness is due to the infill walls,
otherwise it a frequency closer to the previous modes would be expectable. The frequency of
the mode was not as clear as the previous two modes in the FRF but still visible, while the
mode shape presents some incoherence. As it was stated above, the openings in the infills are
not symmetric, which leads to a deviation of the centre of mass from the centre of stiffness
[7], and in the present case it would be expectable that the centre of stiffness would be closer
to the Southeast corner. The mode experimentally detected presents a rotation around a point
closer to the Southwest corner. Similar problems were found in the detected torsional modes
of Models 2 and 3, and, hence, the problem can be associated to an undesirable interaction
between the model and the shaking table, as discussed later. The interaction between the
shaking table and the model is due to the construction process adopted, the transportation
method and the manual bolting of the model to the table. This means that it is impossible to
control possible geometric irregularities of the foundation RC ring beam and the connection
of the building with the shaking table is made with a series of springs (steel bars), with the
adjustment in the bolts allowing for small movements and closing gaps in compression.
SERIES 227887 MASONRY ENCLOSURES Project
14
The fourth and fifth detected modes were, respectively, the second longitudinal at 32.84Hz
and the second transverse at 39.43Hz. These mode shapes are characterized by the movement
of the first floor and roof slabs in the same direction but in opposite ways, with inversion of
curvature. Once again, the influence of the infills can be noted by a small component in the
perpendicular direction of the mode. The change in order of these two second modes, when
compared to the first ones, is possibly due to the large stiffness of infill walls of the South
façade (without openings). The FRF functions also present clear peaks for these last two
modes.
The repetition of the dynamic identification tests after each test stage, DI1 to DI3, allowed for
the detection of the decrement of the frequency of all peaks in the FRF that represent the
above mentioned mode shapes, see Figure 2.10 (a). The increase or decrease of the gain
factor along the dynamic identifications can overlap nearby peaks of other mode shapes,
hence the changes in the FRF need to be tracked in more than one output signal. The damage
in the structure does not only affect the value of the frequency but the shape of the mode as
well, and it is possible, with a considerable amount of damage, for the mode shapes to
disappear, merge or change order. In order to track the evolution of the mode, ensuring a
correct comparison along the loading stages, the Model Assurance Criterion (MAC) [10] was
used:
,
∑ ∅ ∅
∑ ∅ ∑ ∅ (1)
where ∅ and ∅ are the eigenvectors for two different dynamic identification tests and is
the number of degrees of freedom. The MAC was used to compare each mode shape,
identified from DI1 to DI3, with the mode shapes identified in DI0 and it ranges from 0 (no
correlation) to 1 (perfect correlation).
SERIES 227887 MASONRY ENCLOSURES Project
15
1st Transverse Mode
(7.71Hz)
1st Longitudinal Mode
(9.62Hz)
Torsional Mode
(26.95Hz)
2nd Longitudinal Mode
(32.84Hz)
2nd Transverse Mode
(39.43Hz)
Figure 2.9: Mode shapes of the DI 0 of model 1 (initial dynamic identification test)
(a) (b)
Figure 2.10: Frequency change along loading stages: (a) variation of the FRF’s along the test of model 1 at the accelerometer BNE – 1T; (b) evolution of the frequencies along the test of model 1 and their final
variation in respect to DI 0
Figure 2.10 (b) presents the frequency variation of the identified mode shapes along the
dynamic identifications. All five mode shapes were identified from DI0 to DI2, while on DI3
the first two modes, 1st transverse and 1st longitudinal, merged into a single mode shape due to
2 4 6 8 101
2
3
4
5
6
6.41 Hz
7.28 Hz
Gai
n F
acto
r
Frequency (Hz)
DI 0 DI 1 DI 2 DI 3
1st Transversal
7.71 Hz
DI 0 DI 1 DI 2 DI 35
10
15
20
25
30
35
40
45
27.82 Hz(29.5%)
20.88 Hz(36.4%)
39.43 Hz
32.84 Hz
6.43 Hz(16.9% - 33.4%)
9.62 Hz
7.71 Hz
17.42 Hz(35.4%)
Freq
uenc
y (H
z)
Dynamic identification
1st Transversal
1st Longitudinal Torsion
2nd Longitudinal
2nd Transversal
26.95 Hz
SERIES 227887 MASONRY ENCLOSURES Project
16
damage in the RC structure. After the first two stages of the shaking table test, all the
identified mode shapes had and average frequency decrease, in regard to DI0, of 3% and the
first three modes had an average MAC of 0.934. This means that the RC structure was barely
damaged after the 275YRP and 475YRP seismic actions, stages 1 and 2 respectively, and that
the first three mode shapes remained unaltered. This is in agreement with the observed results
since the structure did not present any visible damage after these first two stages. The fourth
and fifth mode presented a lower average MAC of 0.559. Given the small frequency variation
and subsequent lack of damage, this low value can be associated to the difficulties of
capturing more complex mode shapes.
After the third stage, the average frequency decrease of all modes, in respect to DI0, was
30.3% and the average MAC of the last three modes was 0.390. The first two modes merged
into a single mode with a diagonal translation following the Southeast – Northwest direction.
These results seems not in full agreement with the recorded damage after stage 3 (2475 YRP),
see Figure 2.5 and Figure 2.6, which is not enough to assume a loss of almost one third of the
total stiffness of the structure. On the other hand, the collapse of the structure during the last
stage is in agreement with the dynamic data since the structure was already considerably
damaged.
Table 2.1 presents the experimental estimation of the damping ratios along the several
dynamic identifications. None of the identified mode shapes had the expected damping
increment along the tests, confirming the difficulties in the experimental estimation of this
parameter.
Table 2.1 - Experimental damping ratio of model 1
1st
Transverse 1st
Longitudinal Torsion
2nd Longitudinal
2nd Transverse
DI 0 (%) 14.56 3.46 2.15 2.25 0.54
DI 1 (%) 15.44 3.26 2.76 2.05 -
DI 2 (%) 9.98 2.99 2.20 1.81 0.89
DI 3 (%) 4.00 5.33 3.60 4.03 4.00
The seismic vulnerability curves presented in Figure 2.11 relate the damage indicator , see
(2, with the PGA recorded at the base of the model and the computed Input Energy, see (8, for
each mode shape. The damage indicator is computed as:
SERIES 227887 MASONRY ENCLOSURES Project
17
1 (2)
where is the damage indicator of a given mode at stage , is the frequency of the given
mode at stage and is the undamaged, or initial frequency of the given mode. This linearly
proportional ratio between any frequency and the first frequency (DI0), varies from 0,
representing an undamaged state, to 1, representing the collapse of the structure. The damage
indicator assumes isotropic damage [20] between DI0 and stage . The damage indicator of
the torsional mode was associated to the direction with the highest recorded PGA and Input
Energy, hence the longitudinal direction in case of model 1.
The damage indicator is in agreement with the observed damaged, with a very low value after
the first two stages (225 and 475 YRP) and a considerable leap after the third stage (2475
YRP). With the exception of the 1st transverse mode, all other modes have a damage indicator
between 0.30 and 0.36 after the third stage, confirming a generalized loss of stiffness of the
structure and the evenly distributed damage along the four façades of the structure that was
observed. With the collapse of the structure during stage 4 (4574 YRP) along the transverse
direction, the damage indicator of the transverse modes reached the unitary value for the
maximum recorded PGA at that stage.
Figure 2.11: Seismic vulnerability curves of model 1 in the transverse and longitudinal directions, using
the PGA and the Input Energy as input. Here, the damage indicator is a measure of the frequency change
0 1 2 3 4 5 6 7 8 9 100.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Dam
age
indi
cato
r d
PGA (m/s2)
1st Transversal
2nd Transversal
0 1 2 3 4 50.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Dam
age
indi
cato
r d
Input Energy (J)
1st Transversal
2nd Transversal
0 1 2 3 4 5 6 7 8 9 100.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Dam
age
indi
cato
r d
PGA (m/s2)
1st Longitudinal Torsional
2nd Longitudinal
0 1 2 3 4 50.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
Dam
age
indi
cato
r d
Input Energy (J)
1st Longitudinal Torsional
2nd Longitudinal
SERIES 227887 MASONRY ENCLOSURES Project
18
Modal frequencies of the infill walls
Following the same procedure used for the global mode shape identification, the peak
identification in the FRF’s, the frequencies of the first mode shape of the infill walls, in the
North and South façades, were identified in the first dynamic identification, DI0, and their
evolution was registered in the subsequent dynamic identifications, see Figure 2.12. Only the
infills of the North and South were identified due to the fact that the used accelerometers
could not read any information above 80Hz, and the preliminary estimations showed that the
infills of the East and West façades, which are considerably smaller in length when compared
to North and South ones, had the first mode shape at higher frequencies.
The infills of model 1 had two leaves, an exterior one with a 9cm thickness and an interior
one with a 7cm thickness. The results showed that the exterior leaves have a slightly higher
frequency when compared to interior ones, which is expected because the stiffness increases
to the third power of the thickness while the mass only increases linearly. The reason for the
small increase is likely to be the boundary conditions, as the exterior leaves are partly
overhanging the slab, thus with lower restriction to rotation. The infills of the South façade
have a higher frequency than the infills at the North façade, in the same position, due to the
lack of openings. The infills at the second floor have a higher frequency than the ones at the
ground floor.
After the first test stage, DI1, the infill walls did not present any considerable frequency
decrease, in accordance with the observed damage and the dynamic information of the global
structure. On the other hand, after the second stage, when no damaged was observed and no
considerable frequency decrease was registered in the global structure, the infill walls of the
ground floor of the south façade and the exterior leaf of the ground floor of the North façade
presented a frequency decrease of 16.4%, 7.7% and 4.2%, respectively. This frequency loss,
since the walls did not present any visible damage, is likely to be due to the loss of connection
between the infill and RC frame, which makes the wall more flexible. The in-plane damage of
the infills is associated to the interstorey drifts, and in stage 2 a 5.9 mm displacement,
corresponding to 0.30% drift, was recorded at the ground RC frames in the transverse
direction, see Figure 2.13, hence the loss of connection between the infill wall and the RC
frame.
After stage 3, the infill walls of the South façade had an average frequency loss of 16.4%
while the walls on the North façade had an average frequency loss of 15.0%. In the South
façade, the exterior and interior leaves, both in the ground and first floors, converged to the
same frequency after in DI3, which indicates larger damage in the exterior walls. The same
SERIES 227887 MASONRY ENCLOSURES Project
19
situation was not registered in the North façade, where stiffness reduction was proportional,
with the exception of the P1 external leaf.
The infill walls of the ground floor presented a higher frequency loss when compared to the
ones on the first floor, which is in agreement with the observed damage but not with the
interstorey displacements and drifts, since the first storey registered similar or higher values.
Also, none of the infill walls on the first storey collapsed during stage 4, apart from an
exterior jamb on the East façade, while all the infill walls on the ground floor collapsed. The
exterior leaf of the infill wall of the ground floor at the North façade presented a frequency
loss of 43.1%, which is in agreement with the observed damaged since this infill was more
damaged than any other in the transverse direction, and it was one of the first walls to
completely collapse out-of-plane.
North façade infill walls South façade infill walls
Figure 2.12: Evolution of the frequencies of the infill walls in the North and South façades along the test of model 1 and their final variation in respect to DI 0.
Interstorey displacements and drifts
Figure 2.13 presents the interstorey displacements and drifts in each main direction,
transverse and longitudinal, for the three first test stages. Increasingly higher displacements
were recorded for each test stage, as expected, with the exception of the displacements in the
transverse direction in stage 3 (2475 YRP). In the first stage (225 YRP), both directions
presented a similar behaviour and similar maximum displacement values, while on the second
stage the transverse direction was considerably more flexible, with three times larger
displacements than the longitudinal direction. In the third stage, again, both directions have a
similar shape and the maximum displacements are similar. These results are in agreement
with the dynamic identification, since until the second stage (475 YRP) the first mode shape
DI0 DI1 DI2 DI330
35
40
45
50
55
60
65
7064.8 Hz(3.0%)
56.2 Hz(5.5%)
52.7 Hz(8.4%)
34.4 Hz(43.1%)
66.7 Hz
60.5 Hz59.5 Hz
P1 exterior leaf P1 interior leaf P2 exterior leaf P2 interior leaf
Fre
quen
cy (
Hz)
Dynamic identification
57.5 Hz
DI0 DI1 DI2 DI350
55
60
65
70
75
62.3 Hz(7.2%-11.8%)
50.9 Hz(20.0%-26.7%)
70.7 Hz69.4 Hz
67.1 Hz
P1 exterior leaf P1 interior leaf P2 exterior leaf P2 interior leaf
F
requ
ency
(H
z)
Dynamic identification
62.8 Hz
SERIES 227887 MASONRY ENCLOSURES Project
20
is in the transverse direction while the second one is in the longitudinal direction. After the
third stage these two modes merged into a single mode shape which has a diagonal
translation, due to a similar stiffness in both main directions.
In the transverse direction, the first storey recorded increasingly higher drifts when compared
to the ground storey, while on the longitudinal direction the first storey recorded increasingly
lower drifts than the ground floor. In the first stage, the model presented similar drifts in both
transverse and longitudinal directions, while on the second stage, just as in the maximum
displacements, the transverse direction presented significantly higher drifts. The decrement of
the modal frequencies of the infills in the North and South façades only correlates with the
0.30% drift recorded at the ground level but not with the 0.54% drift recorded at the second
floor in the transverse direction during stage 2. After the third stage the drift values of both
main directions are very similar with a 15.6% difference, on average, between them.
Transverse direction Longitudinal direction
Figure 2.13: Interstorey displacements and drifts of model 1
PGA of the infill walls and RC structure
As damage increases along the test stages, the RC structure and the infill walls lose stiffness
but there seems to be no clear trend with respect to amplifications of the base accelerations.
Figure 2.14 presents the maximum recorded acceleration at the infills, in any of the
0 1 2 3 4 5 6 7
stage 1 stage 2 stage 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
00 1 2 3 4 5 6 7 8
stage 1 stage 2 stage 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6
stage 1 stage 2 stage 3
0
storey 1(2 meters)
Drift (%)
0.00 0.05 0.10 0.15 0.20 0.25 0.30
stage 1 stage 2 stage 3
storey 1(2 meters)
Drift (%)
roof(4 meters)
0
SERIES 227887 MASONRY ENCLOSURES Project
21
accelerometers placed at the infill wall, and at the slab levels of the first storey and roof for
each test stage, as well as the maximum amplification, obtained by dividing the maximum
acceleration by the PGA of that same direction. Analysing the RC structure, as expected, the
measured accelerations increased along the test stages in all directions and floors. As for the
amplifications in the longitudinal direction, apart from the roof level from stage 1 (225 YRP)
to stage 2 (475 YRP), the recorded values increased slightly along the three test stages, while
on the transverse direction the amplifications decreased from stage 2 to stage 3 (2475 YRP) at
both levels. The lower stiffness of the transverse direction is in agreement with the observed
collapse mode during stage 4 (4574 YRP), see Figure 2.7. On average, it can be said that the
amplification in the RC structure is not significant.
The infill walls, on the first and second stage, presented similar maximum values for the same
façade, with the exception of the outer leaf at the ground level in the North façade, which
exhibited higher values than the other walls. During the third stage, the maximum recorded
accelerations were no longer similar between the walls of the same façade, but no particular
pattern regarding the position or leaf was found. Very similar values were also recorded in the
infill walls of the same direction, North-South and East-West, during stages 1 and 2. In stage
3, higher maximum accelerations were recorded on the North and East infill walls when
compared to the South and West ones, respectively.
The infill walls of the North and South façades presented a small amplification decrease from
the stage 1 to stage 2, except for the interior leaf at the ground floor at the North façade which
presented a 24.0% decrease and another 15.6% decrease from stage 2 to stage 3. All other
infill walls in the North façade had an amplification increment from stage 2 to stage 3, while
on the South façade the exterior leaf at the ground level and the interior leaf in the first storey
presented a small decrease in the amplification while the outer leaves presented an increment.
In the East and West façades all the infill walls presented a small amplification increment
from stage 1 to stage 2, while on stage 3 half presented a small increment and the other half a
small decrement in the recorded amplification, without any particular pattern as far as the
level or leaf are concerned. Therefore, no clear conclusion can be made regarding the
decrease or increase of amplification with damage, with different trends found.
SERIES 227887 MASONRY ENCLOSURES Project
22
North façade South façade
East façade West façade
RC structure
Figure 2.14: Recorded PGA and amplifications at the infill walls and at the RC structure for each test stage of model 1
Out-of-plane PGD and deformation of the infill walls
Figure 2.15 presents the maximum out-of-plane displacement (PGD), based on double
integration of the measured maximum acceleration, of each infill wall at the North and South
façades and the recorded displacement on the other two accelerometers at the same instant.
Therefore, the line in the graphs of Figure 2.15 represents the out-of-plane deformation of the
infill wall along its length seen from above, and the opposite curvatures of the lines are
associated to interior or exterior bending. It is stressed that these are the displacements of the
5 4 3 2 1 00 5 10 15 20 25 30
1
2
3
NE P2 NE P1 NI P2 NI P1
Amplification Acceleration (m/s2)
Stage
2.0 1.5 1.0 0.5 0.00.0 2.5 5.0 7.5 10.0 12.5 15.0
1
2
3
SE P2 SE P1 SI P2 SI P1
Amplification Acceleration (m/s2)
Stage
2.0 1.5 1.0 0.5 0.00.0 2.5 5.0 7.5 10.0 12.5 15.0
1
2
3
EE 2.1 EE 2.2 EE 1.1 EE 1.2 EI 2.1 EI 2.2 EI 1.1 EI 1.2
Amplification
Stage
Acceleration (m/s2)
1.5 1.0 0.5 0.00.0 2.5 5.0 7.5 10.0
1
2
3
WE 1.1 WE 1.2 WI 1.1 WI 1.2
Amplification
Stage
Acceleration (m/s2)
2.0 1.5 1.0 0.5 0.00.0 2.5 5.0 7.5 10.0 12.5 15.0
1
2
3
Storey 1 Trans Dir Storey 1 Long Dir Roof Trans Dir Roof Long Dir
Amplification Acceleration (m/s2)
Stage
SERIES 227887 MASONRY ENCLOSURES Project
23
infill with respect to the frame, being therefore a relative displacement and not a full
displacement.
As expected, the deformation increases along the seismic tests and the infill walls of the first
storey have a higher deformation than the ones at the ground floor. With the exception of the
infill walls on the first storey at the North façade, the exterior leaf presents a higher
deformation when compared to the interior one. The exterior leaf has a higher thickness than
the interior one, 9cm and 7cm respectively, but the exterior leaf is partially out of the plane of
the RC frame. As noted before, this confirms the fact that the less restraint to rotation is found
on the on the exterior walls, when compared to the interior walls. Even though the infill walls
in the South façade do not have openings and the ones on the North façade have openings, the
range of displacements along the wall is the same for both façades, indicating a small
influence of the openings.
The deformation shape of the infill wall at the ground floor of the North façade was not
altered with the damage along the seismic tests, and both leaves presented the same shape
with PGD recorded at the middle of the wall. Only on stage 3 (2475 YRP) did the
accelerometer next to the door of the exterior leaf, one of the most damaged infills of
model 1, presented a displacement closer to the middle one, confirming the important damage
found. The interior and exterior leaves of the infill wall at the first storey presented the same
deformation shape, but with maximum values at opposite sides under the windows, along the
three recorded test stages.
The interior leaf of the infill wall at the ground floor of the South façade did not present any
visual damage, even though its modal frequency decreased 26.7% from DI0 to DI3, and its
deformation shape did not change along the tests with all three accelerometers recording
similar displacements, hence confirming that the loss of stiffness is related to the loss of
connection between the infill wall and the RC frame. A similar situation can be found in the
exterior leaf until the second stage (475 YRP), since in the third stage (2475 YRP) some
damage was recorded near the RC columns and the displacements at the side and centre
increased. At the first storey, both leaves maintained their deformed shape along the test, and
once again no visual damage was recorded along the test but a decrement in the modal
frequencies was, and nearly the same PGD at the centre and East side of the wall, while on the
West side the exterior leaf had increasingly higher values.
SERIES 227887 MASONRY ENCLOSURES Project
24
North walls South walls S
tage
1
(225
YR
P)
Sta
ge 2
(475
YR
P)
Sta
ge 3
(247
5 Y
RP
)
Figure 2.15: Out-of-plane deformation of the North and South infill walls along the tests of model 1
Figure 2.16 presents the PGD recorded in the infill walls of the East and West façade for the
first three test stages of model 1. As expected, the PGD increases along the test stages and
higher values were recorded in the infill walls of the first storey. The interior leaves of the
infills at the West façade presented the lowest PGD, while all the other infill walls presented
similar values, which is in agreement with the observed damage since both façades presented
a similar crack pattern.
Acc 1 Acc 2 Acc 32.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Out
-of-
plan
e di
spla
cem
ent (
mm
) N-E-P2 N-E-P1 N-I-P2 N-I-P1
Acc 1 Acc 2 Acc 32.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Out
-of-
plan
e di
spla
cem
ent (
mm
)
S-E-P2 S-E-P1 S-I-P2 S-I-P1
Acc 1 Acc 2 Acc 34.04.55.05.56.06.57.07.58.08.59.09.5
10.0
Out
-of-
plan
e di
spla
cem
ent (
mm
)
N-E-P2 N-E-P1 N-I-P2 N-I-P1
Acc 1 Acc 2 Acc 34.04.55.05.56.06.57.07.58.08.59.09.5
10.0
Out
-of-
plan
e di
spla
cem
ent (
mm
)
S-E-P2 S-E-P1 S-I-P2 S-I-P1
Acc 1 Acc 2 Acc 312
14
16
18
20
22
24
26
28
30
Out
-of-
plan
e di
spla
cem
ents
(m
m)
N-E-P2 N-E-P1 N-I-P2 N-I-P1
Acc 1 Acc 2 Acc 312
14
16
18
20
22
24
26
28
30
Out
-of-
plan
e di
spla
cem
ents
(m
m)
S-E-P2 S-E-P1 S-I-P2 S-I-P1
SERIES 227887 MASONRY ENCLOSURES Project
25
Figure 2.16: Out-of-plane PGD of the East and West infill walls of model 1
2.2.2 Results of model 2
Model 2 was designed following the Eurocode normative, EC2 [29] and EC8 [30], using
concrete and steel rebars of higher classes (C30/37 and S500, respectively), together with
single leaf clay brick infill walls with bed joint reinforcement, connected to the RC frame
with steel dowels (or connectors), every second bed joints. Therefore, model 2 represents
likely future solutions for RC frames and masonry infills, where the reason to use a single leaf
is to place an external thermal insulation system. Figure 2.17 presents the position, and label,
of the accelerometers in model 2, noting that the accelerometers were placed on the exterior
only, since walls are single leaf.
North South East West
Figure 2.17: Position and label of the accelerometers in model 2
Overall damage and crack patterns
Model 2 was tested following the same test procedure as model 1 in four stages with
increasing seismic amplitude and, as before, no relevant damage due to transportation
occurred. After the first two stages (225 and 475 YRP), and again as the previous model,
0 5 10 15 20 25 30 35 40
1
2
3
EE 2.1 EE 2.2 EE 1.1 EE 1.2 EI 2.1 EI 2.2 EI 1.1 EI 1.2
Stag
e
Out-of-plane PGD (mm)
0 5 10 15 20 25
1
2
3
WE 1.1 WE 1.2 WI 1.1 WI 1.2
Stag
e
Out-of-plane PGd (mm)
BNE 2L
BNE 1L
N1 4
N1 1N1 3
N1 2
INP L
N1 5
BSW 2L
S2 6
S2 9
S2 3
S2 7
S2 1S2 4
BSW 1L
S1 4
S1 7
S1 1S1 6S1 3
S1 9
S2 2
S2 8
S2 5
S1 8
S1 2
S1 5E1.1 2
E1.1 3
E1.2 1
E2.1 3
E1.1 1
E2.1 1
E2.1 2
BNE 2T
BNE 1T
E2.2 2E2.2 1
INP T
BSW 2T
BSW 1T
W1.2 1W1.1 1
SERIES 227887 MASONRY ENCLOSURES Project
26
model 2 did not present any visible damage but, on the contrary to model 1, model 2
presented negligible frequency decrease, hence negligible stiffness loss. After stage 3 (2475
YRP), the model presented the crack pattern shown in Figure 2.18, with all damage
concentrated at the ground floor. The concentration of lines around the RC columns
represents mortar rendering expulsion, leaving nearly half of the RC column visible, although
no cracks were visible in the RC. Cracks starting from the corners of the openings and
progressing towards the RC frame were also visible after stage 3 in most of the openings. In
the East and West façade, the crack pattern around several jambs is clear, separating them
from the RC frame and the section of the infill wall below the opening, just as in model 1 but
not as clear. The inside face of the model also presented expulsion of the rendering at the
intermediate columns of the East and West façades, leaving the RC columns visible, and a
crack pattern similar to the outside one. After stage 3, the RC structure and the infill walls of
the North and South façades presented an average frequency loss in the identified modes of
28.1% and 17.8%, respectively.
Model 2, contrary to model 1, did not collapse during the fourth and last stage of the test
(4574 YRP), but it was heavily damaged as shown in Figure 2.19 and Figure 2.20. The South
façade, see Figure 2.20 (b), presented the lowest amount of damage, with the first level infill
presenting no cracks within the wall and the ground level infill presenting cracks mainly at
the connection between the infill wall and the RC frame. All the mortar applied to the first
floor columns and part of the mortar of the second floor columns was expelled. On the North
façade, see Figure 2.20 (a), the first storey infill presented cracks at the lateral and upper
connections of the infill to the RC frame, at the intermediate jamb and cracks starting at the
corners of the openings moving towards the RC frame. The ground infill was the most
damaged in the model, as it became completely detached from the surrounding RC frame and
was prevented from falling out-of-plane only by the bed joint reinforcement and the
connectors. The intermediate jamb was completely loose and could be hand pushed out-of-
plane. The East and West façades, see Figure 2.20 (c), presented similar damage, with all the
jambs completely detached from the RC frame and the lower part of the infill wall sustained
only by the bed joint reinforcement and connectors to the RC frame.
All RC columns at the ground level, and part of the RC beams and columns of the second
floor, were visible since the mortar rendering was expelled, and heavy damage was visible.
Columns had mid-height horizontal cracks, see Figure 2.20 (d) and (e), due to the influence of
the infill openings on the horizontal load transfer, aligned with the lower part of the window
openings. One of the columns, see Figure 2.20 (f), presented severe cracking at the upper
SERIES 227887 MASONRY ENCLOSURES Project
27
connection to the beam with rebar exposure, meaning that model 2 could be developing a very
undesirable soft storey collapse mode [40], just as the previous model.
North South
East West
Figure 2.18: Crack patterns of model 2 after stage 3 (2475 YRP) (Note: the drawn lines on the RC frame represent damage on the rendering applied to the RC frame)
North South
East West
Figure 2.19: Crack patterns of model 2 after stage 4 (4574 YRP) (Notes: the drawn lines on the RC frame represent damage on the rendering applied to the RC frame. The blue lines developed after stage 3)
SERIES 227887 MASONRY ENCLOSURES Project
28
(a) (b)
(c) (d)
(e) (f)
Figure 2.20: Damage in model 2 after the fourth stage (4574 YRP): (a) North façade; (b) South façade; (c) West façade from the inside; (d) detail of the left jamb of the door on the North façade and a
horizontal crack at mid-height of the Northeast corner column; (e) horizontal crack at mid-height of the Southwest corner column; (f) heavily damaged top column-beam connection of the Southwest corner
column with loss of the concrete cover and rebar exposure
Modal frequencies of the RC structure
During the first dynamic identification, DI0, five mode shapes were found, see Figure 2.21,
namely: first and second order transverse; first and second order longitudinal; (first) torsional.
The modes and the order of the modes were the same as in model 1. The first transverse and
the first longitudinal had close frequencies, 7.32Hz and 8.37Hz respectively, being the
longitudinal stiffness slightly higher than the transverse one. The frequency leap to the
torsional mode, at 26.77Hz, is associated to the contribution of the infill walls to the global
stiffness of the RC structure. The last two modes, second order longitudinal and second order
transverse, were identified at 30.33Hz and 36.40Hz, respectively.
SERIES 227887 MASONRY ENCLOSURES Project
29
As in mode 1, the torsional mode shape was not perfectly identified, as far as the shape is
concerned, since the North façade did not keep a 90º angle with the other two façades, which
is impossible due to the presence of the RC slab. The second order transverse mode also
presented some distortion at the corners, which is again impossible due to the RC slab. These
problems are associated to the limitations of the used equipment and reading errors. The
dynamic identifications had high coherences, close to the unit value, ensuring the good
quality of the results, and the peaks in the FRF’s were also perfectly clear, see Figure 2.22 (a),
along the test. The MAC [10]values were used to better understand the changes in the mode
shapes.
1st Transverse Mode (7.32Hz)
1st Longitudinal Mode (8.37Hz)
Torsional Mode (26.77Hz)
2nd Longitudinal Mode (30.33Hz)
2nd Transverse Mode (36.40Hz)
Figure 2.21: Mode shapes of the DI 0 of model 2 (initial dynamic identification test)
Figure 2.22 (b) presents the frequency variation of the model along the test. After the first two
stages (225 and 475 YRP) the model did not present any significant frequency decrease in the
mode shapes, when compared to DI0. This is in agreement with the observed results as the
model did not present any visible damage after these two stages, and the reinforcement
connectors prevents the masonry infills to separate from the RC frame. During stage 3 (2475
YRP) the model endured considerable damage in the RC structure, with an average frequency
SERIES 227887 MASONRY ENCLOSURES Project
30
loss of 13.0% in the transverse direction (first and second modes) and 38.2% in the
longitudinal direction (first and second modes) when compared to DI0. The first transverse
and longitudinal modes changed positions and the torsion mode also had a frequency decrease
of 36.7%, with an average MAC of 0.81 which indicates that the mode shapes are only being
slightly altered by the damage. These results are due to the damage observed, possibly due to
damage concealed in the RC structure by the mortar rendering and separation between the
infill walls and the RC frame.
The last stage (4574 YRP) left the model at a near collapse state and the first transverse and
the first longitudinal modes merged, with an average loss of frequency of 78%, with the new
mode being a torsion with the centre of rotation very close to the South face, therefore with a
high movement amplitude of the North façade. These results are in clear agreement with the
observed damage, as the RC structure and the infill walls presented heavy damage, and the
North façade presented the most damaged infill wall at the ground floor and a RC column
with heavy damage and exposed rebar at the connection with the beam. Only the second
longitudinal mode was also identified, and it presented a 56.6% frequency decrease.
The seismic vulnerability curves presented in Figure 2.23 confirm the observed damage and
dynamic data, as until stage 2 (475 YRP) none of the mode shapes present significant damage
and after stage 3 (2475 YRP) the longitudinal modes presented an average damage of 0.38
while the transverse modes presents an average damage of 0.14. The crack pattern observed is
more associated to the transverse damage value than the longitudinal one. After stage four
(4574 YRP), the first and second mode presented a damage around 0.8, indicating the already
mentioned near collapse state of the RC structure and of some of the infill walls.
(a) (b)
Figure 2.22: Frequency change along loading stages: (a) variation of the FRF’s along the test of model 2 at the accelerometer BNE – 2T; (b) evolution of the frequencies along the test of model 2 and their final
variation in respect to DI 0
0 2 4 6 8 101
2
3
4
5
6
7
8
9
10
11
7.03 Hz 7.32 Hz
DI 0 DI 1 DI 2 DI 3 DI 4
1.73 Hz
6.35 Hz
Gai
n F
acto
r
Frequency (Hz)
1st Transversal
7.32 Hz
DI 0 DI 1 DI 2 DI 3 DI 40
5
10
15
20
25
30
35
40
20.41 Hz(32.7%)
16.95 Hz(36.7%)
1st Transversal
1st Longitudinal Torsion
2nd Longitudinal
2nd Transversal
36.40 Hz
30.33 Hz
6.35 Hz(13.2%)
4.72 Hz(43.7%)
31.29 Hz(14.0%)
1.73 Hz(76.3%)
1.64 Hz(80.5%)
8.38 Hz
7.32 Hz
15.79 Hz(56.6%)
Fre
quen
cy (
Hz)
Dynamic identification
26.77 Hz
SERIES 227887 MASONRY ENCLOSURES Project
31
Figure 2.23: Seismic vulnerability curves of model 2 in the transverse and longitudinal directions, using the PGA and Input Energy as input
Modal frequencies of the infill walls
Figure 2.24 presents the frequency decrease of the infill walls on South façade and at the
ground floor of the North façade. As expected, the infills of the South façade present a higher
frequency as they have no openings. The infill walls present an initial small frequency
decrease, after stage 2 (475 YRP), even without any visible damage, possibly associated to
some loss of connection between the infill wall and the RC frame. After stage 3 (2475 YRP),
the infill walls at the ground floor presented a frequency decrease of around 20%, while the
infill at the upper floor presented a decrease of 12%. These results are in agreement with the
crack pattern, since the upper floor presented no visual damage but the ground floor did.
After the last stage (4574 YRP) the infill wall at the ground floor of the North façade was so
damaged and detached from the RC frame that it was not possible to identify its first modal
frequency. The infill wall at ground floor of the South façade presented a higher frequency
decrease when compared to upper floor infill wall, which is in agreement with the observed
crack patterns, although its damaged was mainly at the connection with the RC frame.
0 1 2 3 4 5 6 7 8 9 10 11 120.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Dam
age
indi
cato
r d
PGA (m/s2)
1st Transversal
2nd Transversal
0 1 2 3 4 5 6 7 80.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Dam
age
indi
cato
r d
Input Energy (J)
1st Transversal
2nd Transversal
0 1 2 3 4 5 6 7 8 9 10 11 120.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Dam
age
indi
cato
r d
PGA (m/s2)
1st Longitudinal Torsional
2nd Longitudinal
0 1 2 3 4 5 6 7 80.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Dam
age
indi
cato
r d
Input Energy (J)
1st Longitudinal Torsional
2nd Longitudinal
SERIES 227887 MASONRY ENCLOSURES Project
32
Figure 2.24: Evolution of the frequencies of the infill walls in the North and South façades along the test of
model 2 and their final variation in respect to DI 0
Interstorey displacements and drifts
The interstorey displacements increased with the seismic amplitude, see Figure 2.25, and low
and similar values were recorded for stage 1 and 2 (225 and 475 YRP) in both transverse and
longitudinal directions. During stage 3 (2475 YRP), the longitudinal direction recorded on
average 40.9% higher displacements, confirming the higher loss of stiffness expressed in the
dynamic identification when compared to the transverse direction. On the last stage of the test
(4574 YRP), the highest displacement was recorded in the transverse direction at the first
floor level, while the roof level recorded higher displacements in the transverse direction. The
displacements recorded during the last stage were, on average for both directions, 85.4%
higher when compared to stage 3.
Model 2 recorded maximum interstorey drifts, see Figure 2.25, below 0.05% and 0.08% in the
first two stages (225 and 475 YRP), respectively, and the longitudinal direction recorded
higher values when compared to the transverse direction. Given the higher stiffness of the
longitudinal direction until the third stage, it would be expectable for the transverse direction
to have higher drifts. On stage 3 (2475 YRP), the recorded values on the longitudinal
direction were 38.7% higher, which is in agreement with the dynamic data, as the first
transverse and first longitudinal modes changed order. During the last stage (4574 YRP), the
amplitude of the maximum recorded drifts was considerably higher than the stage 3 values, on
average 88.6%. Except for the transverse direction during stage 1, the interstorey drift values
were always higher at the ground level in comparison to the upper level, which is in perfect
agreement with the crack patterns, as the damage in the RC structure and the infill walls was
DI0 DI1 DI2 DI3 DI420
25
30
35
40
45
50
55
60
65
70
52.1 Hz(22.3%)
57.4 Hz(11.7%)
48.7 Hz(25.1%)
45.9 Hz(19.3%)
23.6 Hz(64.7%)
67.0 Hz65.0 Hz
56.9 Hz
P1 North facade P1 South facade P2 South facade
Fre
quen
cy (
Hz)
Dynamic identification
SERIES 227887 MASONRY ENCLOSURES Project
33
concentrated at the ground floor. It is noted that values up to 4% were found, which are much
larger than values normally accepted in static tests.
Transverse direction Longitudinal direction
Figure 2.25: Interstorey displacements and drifts of model 2
PGA of the infill walls and RC structure
As expected, the RC structure exhibited increasingly higher accelerations along the test, with
the longitudinal direction presenting larger values than the transverse direction, and the roof
recorded higher values in comparison to the first floor RC slab, see Figure 2.26. The
amplification increased from stage 1 to stage 2 and decreased from stage 2 to stage 3 in the
transverse direction, while the opposite was registered in the longitudinal direction, being kept
constant in the last stage. This confirms the rather complex behaviour of the RC frame /
masonry infill system.
The infill walls on the transverse direction recorded lower accelerations along the test when
compared to the infill walls on the longitudinal direction. On the longitudinal direction, the
infill walls at the first floor always had a PGA higher than the ones at the ground floor, but on
the transverse direction, in stage 3 and 4, the three highest PGA’s were recorded at the ground
infill walls. As for the amplifications, on the transverse direction, the changes were in general
small, with the exception of two walls for the last stage (W 1.1 and E 1.2). For the
0 10 20 30 40 50 60 70 80
stage 1 stage 2 stage 3 stage 4
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0 10 20 30 40 50 60
storey 1(2 meters)
stage 1 stage 2 stage 3 stage 4
Displacement (mm)
roof(4 meters)
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
stage 1 stage 2 stage 3 stage 4
0
storey 1(2 meters)
roof(4 meters)
Drift (%)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
stage 1 stage 2 stage 3 stage 4
storey 1(2 meters)
Drift (%)
roof(4 meters)
0
SERIES 227887 MASONRY ENCLOSURES Project
34
longitudinal direction, the amplifications were larger, with a clear increase from stage 2, and
an enormous amplification for S2 at stage 4 (about 3.5).
Transverse direction Longitudinal direction
RC structure
Figure 2.26: Recorded PGA and amplifications at the infill walls and at the RC structure for each test stage of model 2
Out-of plane PGD and deformation of the infill walls
The out-of-plane deformation of the infill walls was computed as the value of the
displacement of all accelerometers or LED’s, depending on the wall, at the instant when the
maximum PGD was measured in each of the test stages. The results of the last test stage
(4574 YRP) are not presented in the South façade because the violence of the seismic action
led to incoherent and unreliable results in a great number of accelerometers. As expected, all
infill walls presented increasing displacements along the seismic test due to the higher seismic
amplitude and higher flexibility, as a result of the accumulated damage, and in the South
façade the infill wall of the first storey presented higher displacements than the ground level
one, see Figure 2.27 and Figure 2.28. Until the third stage (2475 YRP), the infill wall on the
first storey of the South façade did not present any change in the deformed shape, with the
highest values recorded at upper part of the infill, which is in agreement with the observed
crack pattern since the infill did not present any visible damage and the frequency loss is
associated to damage in the connection to the RC frame. The infill wall at the ground floor
2 1 00 5 10 15 20
1
2
3
4
E 1.1 E 1.2 E 2.1 E 2.2 W 1.1 W 1.2
Amplification
Stage
Acceleration (m/s2)
5 4 3 2 1 00 5 10 15 20 25 30 35 40
1
2
3
4
N1 S1 S2
Amplification Acceleration (m/s2)
Stage
2 1 00 5 10 15 20
1
2
3
4
Storey 1 Trans Dir Storey 1 Long Dir Roof Trans Dir Roof Long Dir
Amplification Acceleration (m/s2)
Stage
SERIES 227887 MASONRY ENCLOSURES Project
35
presented the highest displacements, until the second stage (475 YRP), at the right side of the
wall where more extensive cracks were observed and in the third stage (2475 YRP) also at the
top. This change is associated to the observed damage.
The infill wall at the first floor of the North façade, see Figure 2.29, presented its highest
displacements below the left window until the third stage (2475 YRP) of the test, and the lack
of change in the deformed shape is associated to the absence of localized damage, as the
increment in the flexibility is associated to the loss of connection to the RC frame. On the last
stage (4574 YRP), the maximum values were recorded at both sides of the wall. On the South
façade the results were obtained from the double integration of the data recorded by the
accelerometers, while on the North façade the results were obtained using the KRYPTON
camera, and given the difference characteristics, precision and acquisition frequency of the
equipment, the obtained displacements are different and should not be directly compared.
The PGD of all other infill walls on model 2 are presented in Figure 2.30, but only for the
three first stages of the test because some of the results of the last stage were unreliable. All
the infills recorded higher PGD’s along the test and the infills at the first floor recorded higher
PDG in all three stages, when compared to ground floor ones. At the ground floor, the infill in
the North façade recorded the highest PGD in stages 2 and 3, as expected being the lengthiest
in this comparison, and the most damaged in the model.
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2425 YRP)
Figure 2.27: Out-of-plane deformation of the infill wall at the ground level of the South façade (mm)
0.7665 0.7665
0.7010
0.8320
0.6355
0.89750.9630
0.8320
1.029
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
1.3991.399 1.274
1.5231.647
1.7721.896 2.021
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
3.295
3.8134.330
4.848
3.295
5.365
2.777
5.883
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
SERIES 227887 MASONRY ENCLOSURES Project
36
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 2.28: Out-of-plane deformation of the infill wall at the first storey of the South façade (mm)
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP) Stage 4 (4574 YRP)
Figure 2.29: Out-of-plane deformation of the infill wall at the first storey of the North façade (mm)
4.919
4.991 5.064
5.136 5.209
4.846
5.281
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
8.178
8.289
8.4008.511
8.622
8.067
8.733
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)25.32
25.71 26.1026.49
24.9324.54
26.8826.88
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
-0.018 0.120.26
0.12
0.39 0.12
-0.018
0.530.53
-0.018
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
0.480.39
0.29
0.58
0.19 0.098
0.580.67
0.29
0.67
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
0.97
0.44
1.52.0
-0.090
2.63.1
0.97
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
-13-13
-6.30.40 7.1 14 20
-20
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
SERIES 227887 MASONRY ENCLOSURES Project
37
Figure 2.30: Out-of-plane PGD of the North, East and West infill walls (mm)
2.3 DESIGN AND CONSTRUCTION OF THE MODELS
2.3.1 Building model
The first step in the experimental program was the definition of the geometry and the building
solutions of the prototypes. The geometry survey was done elsewhere to define the average
height and length of the RC frames [34], and the resulting geometry, a building with a two
storey single bay frame in one direction and a two storey double bay frame in the other
direction, can be seen in Figure 2.31. Given that one of the objectives of the present work is to
assess the performance of modern RC frame structures, the RC frame was designed according
to the most recent standards, EC2 [29] and EC8 [29], including reinforced solutions for the
infill walls.
Figure 2.31: Prototype geometry (m)
0 5 10 15 20 25 30 35
1
2
3
E 1.1 E 1.2 E 2.1 E 2.2 W 1.1 W 1.2 N 1
Stag
e
Out-of-plane PGD (mm)
Facade A Facade B
B
A5,7
33
3,23 3,23
Plan view
6,45
5,7
SERIES 227887 MASONRY ENCLOSURES Project
38
Testing the complete structure and not only the RC frames with infills seems a more
reasonable option as the interplay of the response of all components can be captured. Ideally,
the test should be performed at full scale but the physical limitations (maximum dimensions
and payload capacity) of the testing equipment and laboratories impose, in most cases, the use
of scaled models in shaking table tests. This is an important drawback of this type of facility.
Scaled models are obtained using similitude laws, which add complexity to the construction
and test setup in order to fulfil their requirements. In order to test a RC model to its ultimate
capacity the following need to be correctly simulated: i) the geometry; ii) the stress-strain
relationship of the materials; iii) the mass and gravity forces; iv) the initial conditions and the
boundary conditions [5].
The first condition is easily fulfilled by direct application of a geometric scale, although some
pre-fabricated construction elements may have a limited range of dimensions, as infill clay
units. Very small scales may also represent higher construction challenges. Obtaining
adequate stress-strain relationships of the materials can be a much more complex task since
such a relationship has to be fulfilled throughout different stress or strain levels, rates,
gradients, etc. [2]. For very small scales it is not uncommon to use different materials in the
models. The mass and gravity forces are addressed, respectively, by the Cauchy and Froude
similitude laws [5]. The first one is adequate for phenomena in which the restoring forces are
derived from the stress-strain constitutive relationships and the elastic restoring forces, see
Eq. (3). Froude similitude is adequate for phenomena in which the gravity forces are
important, being the Froude value the ratio between inertia and gravity forces, see Eq. (4).
The use of both laws simultaneously is the obvious choice in order to more accurately
replicate the dynamic behaviour of structures, particularly when strongly non-linear behaviour
is expected. In the present work both laws were taken into consideration in the model
definition, following the relations described in Table 2.2. As for the boundary conditions, the
soil-structure interaction is not considered as the model is fixed to the shaking table using
bolts, meaning that the input signal is directly transferred to the structure.
(3)
(4)
SERIES 227887 MASONRY ENCLOSURES Project
39
Table 2.2 - Cauchy-Froude similitude law
Parameter Scale Factor Parameter Scale Factor
Length (L)
Mass (m)
Modulus of elasticity ( E ) 1
Weight (w)
Specific mass (ρ) Force (F)
Area (A) Moment (M)
Volume (V) Stress (τ) 1
Displacement (d) Strain (ε) 1
Velocity (v) Time (t)
Acceleration (a) 1 Frequency (f)
High resistance class materials were chosen for the concrete and steel rebars (C30/37 and
S500, respectively), while the scale factor was chosen taking into account the physical
limitations of the shaking table of LNEC, see Chapter 3. The models were designed at a
reduced scale of 1:1.5, meaning that reduced loads were applied to a model with the geometry
also reduced, see Figure 2.32, using the similitude law, see Table 2.2 ( 1.5 . The design
loads used in all three models, reduced using the similitude law relations described in
Table 2.2, can be seen in Table 2.3.
Figure 2.32: Geometry of the tested model reduced to a scale of 1:1.5
SERIES 227887 MASONRY ENCLOSURES Project
40
Table 2.3 - Design loads of the models already reduced at scale of 1:1.5
Load Description Load value
Self-weight
Partition walls mortar (1.5 cm) + clay masonry units (11cm) + gypsum (1.5cm)
1.11 KN/m2
Floor slab storey 1 reinforced concrete slab (thickness = 0.12 m) 1.07 KN/m2
Roof slab reinforced concrete slab (thickness = 0.12 m)- 1.07 KN/m2
Infill walls mortar (1.5cm) + clay masonry unit (9cm) + clay masonry unit (7cm) + gypsum (1.0cm)
3.74 KN/m
Infill wall mortar (1.5cm) + clay masonry unit (15 cm) + mortar (1.5cm)
3.00 KN/m
Imposed Load
Storey 1 domestic and residential 1.33 KN/m2
Roof accessible 0.67 KN/m2
The EC8 [29] contemplates seismic design using response spectra for near-field and far-field
earthquakes. In order to define these spectra, one must know the geographical location of the
building and the soil type of the building. The EC8 [29], as an international standard, refers
specific parameters to the National Annex. Therefore, in article NA-3.2.1(2) of the National
Annex of EC8 [29], continental Portugal is divided in a local council-based zoning,
considering six magnitude levels for Type 1 far-field seismic actions (1 to 6) and five
magnitude levels for Type 2 near-field seismic actions (1 to 5).
As for the soil type, EC8 [29] defines in its Table 3.1 seven types of soil, from rock (type A)
to soils with liquefaction characteristics (type S2). The specific parameters associated to each
soil type needed to compute the response spectra can be found in the National Annex of EC8
[29], more specifically in its Tables NA-3.2 and NA-3.3. Here, it is assumed that the models
would be built in Lisbon, zones 1.3 and 2.3, and in a type A rocky soil. The obtained spectra,
reduced following the similitude law, can be seen in Figure 2.33.
SERIES 227887 MASONRY ENCLOSURES Project
41
Figure 2.33: Response spectra after the application of the similitude law, see Table 2.2, according to EC8 [29]
All four façades of the model had infill walls. The South façade was blind while the others
had openings in around 20% of the surface area, see Figure 2.34. The clay brick units used
were only scaled in the thickness, while height and length were the same for all three models
and kept at a 1:1 scale.
0.0 0.5 1.0 1.5 2.0 2.5
0.25
0.50
0.75
1.00
1.25
1.50 Type 1 Type 2
a (m
/s2 )
T (s)
SERIES 227887 MASONRY ENCLOSURES Project
42
(a) (b)
(c) (d)
Figure 2.34: Geometry of the openings in each façade: (a) North façade; (b) West façade; (c) East façade; (d) South façade
The chosen similitude law implies that the specific mass of the prototype and the model are
different, see Table 2.2, as . This problem was solved using two types of
additional steel masses: one applied to the RC structure; and another to the infill walls. The
masses of the RC structure had 82x82x26 cm, weighted around 12 KN each and were bolted
to the slab of the first floor and the slab of the roof. A total number of twelve masses were
used, six in each slab, see Figure 2.35 (a). The masses applied to the infill walls had 15x15x4
cm and weighted around 0.072 KN each. These masses were applied to both sides of the wall,
evenly distributed and bolted in two edges of the plate, see Figure 2.35 (b). Each mass was
bolted to the surface of a single unit, in order not to increase the strength of the masonry joints
or influence the crack pattern. A total number of three hundred and thirty four masses were
used in each model.
SERIES 227887 MASONRY ENCLOSURES Project
43
(a)
(b)
Figure 2.35: Additional steel masses for: (a) RC concrete structure, bolted to the slabs of the 1st floor and roof with 82x82x26 cm and 12KN each; (b) Infill walls, bolted to both sides of the wall with 15x15x4 cm
and 0.072KN each
The model to be tested in the shaking table was built inside the NESDE building and then was
transported to the shaking table using the existing crane. In the present work, the model was
built by a construction company hired specifically for this project using techniques and
workers accustomed to RC and masonry construction. Given these conditions two aspects had
to be considered: the foundation of the model had to be plane, otherwise the model could be
damaged when bolted to the shaking table; lifting eye bolts had to be provided for the
transportation of the model. The first issue was solved by constructing the model on top of a
horizontally aligned platform, see Figure 2.36 (a). The second issue was solved by designing
the models with a RC ring beam with four steel plates with a lifting eye at its corners, see
Figure 2.36 (b). This ring beam was also perforated in order to bolt the model to the shaking
table. Figure 2.36 (c) shows the model already on top of the shaking table, ready for testing.
SERIES 227887 MASONRY ENCLOSURES Project
44
(a)
(b)
(c)
Figure 2.36: Construction of the models: (a) horizontally aligned surface on which the models were constructed; (b) RC ring beam with steel connector with an eye in lift and transport the model to the
shaking table; (c) model 1 on the shaking table before the test
The infills of the model represent another future possibility when the design follows EC2 [29]
and EC8 [29], a single leaf clay brick wall with reinforced rendering nailed to the RC frame
and infill wall on both sides, see Figure 2.37. The leaf was completely within the RC from
plane, mortar rendering was used on both sides of the infill and the units were horizontally
perforated. The mortar used for the bed joints and plaster was pre-batched and with a M5
class.
The reinforcement grid chosen, Bekaert Armanet ϕ1.05mm 12.7x12.7mm, see Figure 2.38
(a), was nailed to the RC frame using a Hilti X-M8H10-37-P8, see Figure 2.38 (b) and (c),
using a gun and Hilti shot powder actuated tools. Similar nails should have been used to nail
the grid to the infill wall but were substituted by the additional masses above mentioned, see
Figure 2.35 (b), as these had to be used already due to the similitude law chosen. Figure 2.38
(d) shows the application of the grid in the inner surface of an infill. In order to simulate the
attachment of the grid to the infill wall with nails, a ring was installed between the mass and
SERIES 227887 MASONRY ENCLOSURES Project
45
the wall, in each bolt, with a contact area similar to the Hilti X-M8H10-37-P8 nail applied in
the RC frame, see Figure 2.38 (e).
(b)
(a) (c)
Figure 2.37: Single leaf clay brick infill walls with reinforced plaster from Model 3 already scaled: (a) spacing of the Hilti X-M8H10-37-P8 connectors along the height of the RC column; (b) detail at the RC
column; (c) detail of the Hilti X-M8H10-37-P8 connectors
20.0
40.0
40.0
40.0
40.0
RC column
Bekaert ArmanetØ1.05mm 12.7x12.7mm
1.5cm M5 mortar
Hilti X-M8H10-37-P8
1.5cm M5 mortar
RC column
Bekaert ArmanetØ1.05mm 12.7x12.7mm 1.5cm M5 mortar
1.5cm M5 mortar Hilti X-M8H10-37-P8
30x20x15cm
Hilti X-M8H10-37-P8
Bekaert ArmanetØ1.05mm 12.7x12.7mm
1.5cm M5 mortar
SERIES 227887 MASONRY ENCLOSURES Project
46
(a)
(b)
(c)
(d)
(e)
Figure 2.38: Construction of the infills of model 3: (a) Bekaert Armanet ϕ1.05mm 12.7x12.7mm; (b) Hilti X-M8H10-37-P8; (c) application of the grid in the outer surface at a corner column; (d) application of the
grid in the inner surface; (e) additional masses with steel rings attached to the infill walls
SERIES 227887 MASONRY ENCLOSURES Project
47
2.3.2 Wall panels
In order to test the four types of masonry enclosures’ solutions described in Chapter 1, the
second part of this transnational access activity comprised the dynamic testing of a closed RC
plane frame with external dimensions of 6.40m by 3.25m. The structural elements of this
frame have cross-sectional dimensions of 0.50m by 0.40m (beams) and 0.40m by 0.40m
(columns). These plane frame specimens were tested simultaneously for in-plane and out-of-
plane dynamic actions, representing the response of a frame panel in the 4th floor of an eight
storey RC building (see Figure 2.39). The columns have a centred prestress which represents
the vertical load from the floors above.
Figure 2.39: Frame panel of typical RC building
The first two masonry enclosure solutions tested were the unreinforced masonry and the one
with horizontal reinforcement between masonry units (Bekaert Murfor RND/Z-5-200).
Afterwards, both masonry infills were demolished and rebuilt using a reinforced mortar
coating. The following pictures (Figure 2.40 to Figure 2.44) show the construction process of
the models.
Figure 2.40: Reinforcement layout for the RC frames
SERIES 227887 MASONRY ENCLOSURES Project
48
Figure 2.41: Prestressing details
Figure 2.42: RC frame ready for infill construction
Figure 2.43: Masonry infill construction with bed joint reinforcement
SERIES 227887 MASONRY ENCLOSURES Project
50
3 The LNEC Earthquake Engineering testing
facility
LNEC owns a large scale experimental facility for seismic testing of structures which is part
of the European Seismic Engineering Research Infrastructures and whose construction was
partially financed by the European Union. The Earthquake Engineering and Structural
Dynamics Division operates this facility, pictured in Figure 3.1, and develops R&D activity in
the fields of Earthquake Engineering and Structural Dynamics.
Figure 3.1: LNEC Earthquake Engineering testing facility (Ferry Borges building)
The experimental activity carried out in the LNEC earthquake engineering testing facility, and
related research, aims at assessing the performance of structures subjected to dynamic and
seismic loadings. The tests are carefully setup in order to simulate on the models the same
conditions as in the real prototypes and measure all the relevant effects necessary for the
SERIES 227887 MASONRY ENCLOSURES Project
51
performance assessment. The instrumentation is calibrated and a record is made of all the
maintenance and interventions made. The preparation and operation of the shaking tables
follows standard protocols to achieve the targets proposed. The analysis process uses the most
advanced techniques in the fields of signal processing, dynamics of structures and earthquake
engineering to achieve optimum results. All these aspects contribute to guarantee that the tests
carried out meet the highest quality standards.
3.1 GENERAL INFORMATION ON THE LABORATORY
Table 3.1 – Name and location of the Laboratory Full Name of the Laboratory Núcleo de Engenharia Sísmica e Dinâmica de Estruturas
Abbreviated Name NESDE/LNEC
Address Av. Brasil, 101 1700-066 Lisbon
Location Lisbon
Country Portugal
Telephone +351218443824/3307
Telefax +351218443035
E-mail/www http://www.lnec.pt/LNEC/DE/NESDE/
Table 3.2 – Name and location of the parent organization Full Name of the Parent Organization Laboratório Nacional de Engenharia Civil
Address Av. Brasil, 101 1700-066 Lisbon
Location Lisbon
Country Portugal
Telephone +351218443000
Telefax
E-mail [email protected]
3.2 THE FACILITY: LNEC-3D SHAKING TABLE
The facility has a large testing room, shown in Figure 3.2, and includes two shaking tables,
one large triaxial and another one smaller uniaxial, and various other equipment for seismic
testing of structures. The triaxial shaking table, pictured in Figure 3.3, is capable of testing
large civil engineering structures subjected to earthquake motions up to collapse.
SERIES 227887 MASONRY ENCLOSURES Project
52
Figure 3.2: LNEC earthquake engineering testing room
Figure 3.3: LNEC-3D shaking table
SERIES 227887 MASONRY ENCLOSURES Project
53
3.3 GENERAL INFORMATION ON THE SHAKING TABLE
Table 3.3 – Name of the LNEC-3D shaking table Full Name of the Shaking Table LNEC TRIAXIAL SHAKING TABLE
Abbreviated Name LNEC-3D
Designer/Manufacturer LNEC and INSTRON
Year of Installation 1995
3.4 SHAKING TABLE DESCRIPTION
Table 3.4 – Type of shaking table
Longitud. X Transverse Y Vertical Z Pitch Roll Yaw
Uniaxial - - - - - -
Biaxial - - - - - -
Multiaxial Y Y Y N/A N/A N/A
Table 3.5 – Characteristics of the Platform
Size (m×m) 4.6 x 5.6 Weight (kN) 392 Material Steel
Type of Actuation Hydraulic
Table 3.6 – Characteristics of the Actuators
Manufacturer Total Force (kN) Number of units/axis
Longitudinal INSTRON 1250 1
Transverse INSTRON 750 2
Vertical INSTRON 375 1
SERIES 227887 MASONRY ENCLOSURES Project
54
Table 3.7 – Shaking table performances
Frequency Range Hz 0.1 – 40.0
Stroke (effective/maximum) Horizontal mmpp 290/400
Vertical mmpp 290/400
Max Velocity (nominal/limit) Horizontal
Transverse cm/s 70.1/121.5
Longitudinal cm/s 42.4/73.5
Vertical cm/s 41.9/72.6
Max Acceleration at bare table Horizontal
Transverse m/s² 18.75
Longitudinal m/s² 31.25
Vertical m/s² 9.38
Yaw Rotation degrees ° N/A
Velocity rad/s N/A
Pitch/Roll Rotation degrees ° N/A
Velocity rad/s N/A
Max Overturning Moment kN×m N/A
Max Specimen Dead Weight kN 392
Max Compensated Dead Weight kN 392
3.5 CHARACTERISTICS OF THE CONTROL SYSTEM
Type of Control Analogue Digital Mixed
Table 3.8 – Characteristics of the analogue part
Manufacturer LNEC
Type LNEC-CTL
Table 3.9 – Characteristics of the digital part
Hardware
Computer Host PC+NI PXI Real Time Controller+4 RIO FPGA Virtex-5
D/A Channels 8 ADC channels, 16 bit
96 configurable digital channels A/D Channels
Software Designer LNEC
Controlled motions Sinusoidal Random Shock Seismic
No. of Controlled Channels 3 3 3 3
No. of Acquisition Channels 6 6 6 6
SERIES 227887 MASONRY ENCLOSURES Project
55
3.6 COMPLEMENTARY FACILITIES
Floating Foundation Hydraulic System Bridge Cranes Capacity
Dimensions (m×m) - Electric Power (kW) 330 No. of Cranes 2
Weight (kN) - Flow Rate (l/min) 690 Max Load (kN) 392
Natural Freq. (Hz) - Pressure (MPa) 20.7 Useful Height (m) 8
SERIES 227887 MASONRY ENCLOSURES Project
56
4 Sensors technical data
The experimental assessment of a model performance under seismic actions imposed on the
shaking table requires the measurement of several different types of physical quantities like
displacements, accelerations, strains and forces. The sensors necessary for this includes
displacement transducers, accelerometers, strain gauges and load cells, all of them available at
the LNEC Earthquake Engineering and Structural Dynamics Division. Below is listed only
the subset of the existing instrumentation that is relevant for the tests carried out in the scope
of the present study.
4.1 DISPLACEMENT TRANSDUCERS
4.1.1 LVDT displacement transducers
RDP Electronics ACT2000C, ACT4000C and ACT6000 inductive displacement transducers,
having work strokes of +/-50mm, +/-100mm and +/-150mm, respectively, can be used for
measuring displacements. Figure 4.1 shows the general aspect of the inductive displacement
transducers available at LNEC, while Table 4.1 shows some of its main characteristics.
a) ACT captive guided
b) ACT series unguided
Figure 4.1: LVDT displacement transducers (source: RDP)
SERIES 227887 MASONRY ENCLOSURES Project
57
Table 4.1 – Characteristics of the RDP displacement transducers
Manufacturer RDP ELECTRONICS (www.rdpe.com)
Models available ACT2000, ACT4000, ACT6000
Stroke +/-50mm (ACT2000), +/-100mm (ACT4000), +/-150mm (ACT6000)
Sensitivity 15mV/V/mm (ACT6000) to 30mV/V/mm (ACT2000)
Energising supply 5Vrms, 5kHz
Linearity deviation 0.08% (ACT2000) to 0.3% (ACT6000)
4.1.2 Hamamatsu optical system
Optical displacement transducers HAMAMATSU C5949 (comprising F50mm lens, sensor
head and LED target) and HAMAMATSU conditioning PSH controllers C2399 (see
Figure 4.2) can be used for measuring 2D displacements on a plane perpendicular to the line
of sight (typically either on a horizontal or vertical plane). Table 4.2 shows some of the main
characteristics of the HAMAMATSU displacement transducers.
Figure 4.2: HAMAMATSU optical 2D displacement transducer
Table 4.2 – Characteristics of the HAMAMATSU displacement transducers
Type Spectral
Response [nm]
Measurement Points [-]
Sampling Frequency
[Hz]
Position Detecting Error [%]
Resolution [-] Error due to light
[%]
C2399-00 1 300
C5949 700 to 1150 1 to 7 300 ±1 1/5000 ±1
SERIES 227887 MASONRY ENCLOSURES Project
58
The cameras used detect the positions of the points concerned along the x- and y- axes with
respect to their absolute reference system [12]. Figure 4.3 shows a typical Hamamatsu system
configuration.
Figure 4.3: Hamamatsu system configuration [12]
4.1.3 Krypton K600 camera
The K600 camera system takes measurements in 3D and consists of three linear CCD
cameras. While two external cameras, present inside the instrument, detect the position y- and
z- coordinates of the point concerned, the central one calculates the x-coordinate. Thus the
position of an infrared LED is calculated by triangulation [19]. Figure 4.4 shows the Krypton
K600 camera mounted on a tripod.
Figure 4.4: The Krypton K600 camera [19]
The overlap area of the three linear CCD-cameras in the camera unit, results in an overlapping
pyramidal volume as shown in Figure 4.5. The top angle of the pyramid is 34° (+17° / -17°).
SERIES 227887 MASONRY ENCLOSURES Project
59
The K600 Krypton Camera offers a 17m3 field of view volume with accuracies up to ±0.1mm
for static and dynamic LED position measurements. According to the rule-of-thumb, the
lateral visibility limit (measured from the symmetry plane of the camera) is half the distance
from the camera [19].
Figure 4.5: Representative measurement volume of a Krypton K600 camera [19]
The field of view is divided into three accuracy zones which are determined based on the
distance from the camera as seen in Figure 4.6. Accuracy of LED dynamic measurements has
been observed to be higher for motions in the X-Y plane of the viewing volume.
Figure 4.6: Accuracy zones of the Krypton K600 camera system [19]
The acquisition laptop communicates with the camera controller to acquire the 3D position
data of the individual LEDs. The camera control unit synchronizes the LEDs with the
acquisition of the camera and serves as an interface between the laptop and the camera unit.
The strober distributes the control signals from the controller to individual LEDs and
translates them into pulse trains causing the infrared LEDs to flash. The strober can be
connected to the strober port of the controller or can be daisy-chained to another strober. The
LED’s are plugged into a strober and then attached to the object that the user wants to
SERIES 227887 MASONRY ENCLOSURES Project
60
measure. The best results are achieved if they are placed at large distances forming large
triangles as seen by the camera.
4.2 ACCELEROMETERS
Two main types of accelerometers are used in the LNEC Earthquake Engineering and
Structural Dynamics Division: Endevco, model 7290-A with variable capacitance
(Figure 4.7), and PCB Piezotronics, model 337A26 (Figure 4.8). The first type is mainly used
on the shaking table while the second one is mainly used on the mock ups. Both are high
frequency accelerometers adequate for the measurement of accelerations in dynamic and
seismic tests; their main characteristics are summarised in Table 4.3 and Table 4.4.
Figure 4.7: Endevco accelerometers
Figure 4.8: PCB Piezotronics accelerometers
Table 4.3 – Characteristics of the Endevco accelerometers
Manufacturer MEGGITT SENSORS (www.endevco.com)
Model 7290A-2 and 7290A-10
Sensitivity (at 100Hz) [mV/g] 1000+/-20 and 200+/-10
Measurement range [g peak] +/-2 and +/-10
Amplitude response at +/-5% [Hz] 0 to 15 and 0 to 500
Transverse sensitivity [% max] 2
Table 4.4 – Characteristics of the PCB accelerometers
Manufacturer PCB Piezotronics (www.pcb.com)
Model 337A26
Sensitivity [mV/g] 100
Measurement range [g peak] 100
Broadband resolution [g rms] 0.0001
Frequency range [Hz] 0.5 to 5000
SERIES 227887 MASONRY ENCLOSURES Project
61
4.3 LOAD CELLS
Two load cells from Instron were used to measure the forces applied to the wall panels. The
load cell is an electric transducer used to measure a force applied on a mechanical or
structural component, through the measurement of an electrical signal that varies due to the
deformation that produces such a force on the component itself. The relief of the mechanical
deformation takes place in an indirect manner, via a reading in mV or V, and subsequently
transformed into the correct unit of measurement.
An individual load cell is constituted of a hollow steel cylinder. Six strain gauges are
connected to the body of the load cell – four in the longitudinal direction and two in the
transverse direction relative to the axis of the cylinder. The strain gauges are all connected in
a full-bridge configuration to amplify the magnitude of the signal. Figure 4.9 shows the load
cells.
3
Figure 4.9: Instron load cells
4.4 ACQUISITION SYSTEM
The acquisition system available at the LNEC Earthquake Engineering and Structural
Dynamics Division comprises the following components:
Table 4.5 – NI PXI controller
NI PXI-8106 CONTROLLER
2.16 GHz Intel Core 2 Duo T7400 dual-core processor
Up to 46% higher performance than the PXI-8105 512 MB (1 x 512 MB DIMM) dual-channel 667 MHz DDR2 RAM standard, 4 GB maximum 10/100/1000 BaseTX (Gigabit) Ethernet, ExpressCard/34 slot, and 4 Hi-Speed USB ports
Integrated hard drive, GPIB, serial, and other peripheral I/O
Windows OS and drivers already installed; hard-drive-based recovery
SERIES 227887 MASONRY ENCLOSURES Project
62
Table 4.6 – NI PXI chassis
NI PXI-1052 CHASSIS
• 4 slots for 3U PXI modules and 8 slots for SCXI modules • Latest chassis technology • AUTO/HIGH fan-speed selector to optimize cooling and acoustics • 0 to 55 °C temperature range • 42 dBA acoustic emissions • Multiplexed operating mode for SCXI
• SCXI high-voltage analog backplane integrated internally
Form Factor PXI Platform, SCXI
PXI Bus Type PXI Hybrid Compatible Operating System / Target Windows, Real-Time
LabVIEW RT Support Yes
Power Supply AC
Number of Slots12
Number of PXI Peripheral Slots4
Maximum System Bandwidth132 MB/s Accepts both 3U PXI and CompactPCI ModulesYes Optional Front or Rear Rack Mountable Yes
Integrated ControllerNo
Remote Power-inhibit Control and Voltage Monitoring Yes Total Available Power 450 W Input Voltage Range 100..240 V Input Frequency Range 50/60 Hz Field-replaceable Power Supply No
SERIES 227887 MASONRY ENCLOSURES Project
63
5 Test setup
5.1 BUILDING MODEL TEST SETUP
The test setup for the building model was already shown in Figure 2.36. After the base slab
was attached to the shaking table the model was thoroughly instrumented in order to
characterize its behaviour during the tests.
The results of a shaking table test can be obtained via: (a) visual record of the damage,
immediately after the test or at a later stage through video and photographic data; (b) data
acquisition equipment attached to specific points of the model. The model was filmed with a
high frequency video recording system and photographed before the first stage and after each
different stage, in order to produce damage maps and their evolution along the experimental
test. All cracks were painted along the test using different colours to clearly record their
evolution.
Different instruments were installed for recording displacements, both absolute and relative,
and accelerations. The acquisition equipment uses sensors to transform physical quantities
(displacements, velocities and accelerations) in electric signals and the most commonly used
sensor in shaking table tests are accelerometers (ACC). In the present case, these are SDOF
(single degree of freedom) systems having an inertial mass that moves proportionally to the
amplitude of the acceleration of a moving body, which is converted into an electrical signal in
the form of voltage [13]. Displacement transducers (LVDT and infrared cameras) are
frequently used as well, although velocities and displacements can be obtained from the
integration and double integration, respectively, of the acceleration signals. There are several
types of ACC and two different ACC were used here: piezoelectric and capacitive, see
Figure 5.1. The main differences between both systems are the power supply, since
piezoelectric ACC need an external power source, and a limited range of 1000Hz in the
capacitive ACC. Piezoelectric ACC are also capable of measuring uniform acceleration
signals.
SERIES 227887 MASONRY ENCLOSURES Project
64
(a) (b) (c)
Figure 5.1: Accelerometers used in the shaking table tests: (a) piezoelectric from PCB [31], [32]; (b) piezoelectric from Wilcoxon [40]; (c) capacitive from Endevco [9]
Regarding the piezoelectric type, three different models from two different manufacturers
were attached to the structure, all sharing the same sensitivity (1000 mV/g±5%) and
measurement range (±5g). The NESDE has also ACC with a lower measurement range and
higher sensitivity that could provide more accurate results during the dynamic identification
tests, see Table 6.1, but due to their limited range these ACC could not be used during the
seismic tests. Switching ACC between dynamic identifications and seismic tests is not
recommended in such a complex test that involves several kilometres of cables to connect all
the ACC to the acquisition equipment, over forty ACC, all the laboratory technicians and at
least one full day. Switching potentiates mistakes and malfunctions of the highly sensitive
equipment in use, and increases exponentially the time needed to perform the test. The
capacitive ACC were pre-installed in the shaking table in each direction (longitudinal,
transverse and vertical). The definition of the instrumentation setup is based on the expected
response of the model to the input, obtained from preliminary studies [20], and the objectives
of the test, meaning that the instrumentation can be divided in two groups: (i) setup to acquire
the out-of-plane behaviour of the infill walls; (ii) setup to acquire the global behaviour of the
RC concrete structure. The out-of-plane behaviour of the infill walls was captured by a set of
ACC distributed in the surface of the wall, see Figure 5.2. The blind walls received nine ACC
each, with one line at mid-height and the other two at half distance to the RC beams and
columns and the infill walls at the East and west façades received ACC at a third of the height
below the opening. The North infill wall at the ground floor received ACC below the opening,
at a third of the height also and at the centre lines between the openings. A total number of
thirty-four ACC were used.
SERIES 227887 MASONRY ENCLOSURES Project
65
(a) (b) Figure 5.2: Accelerometers setup: (a) North and East façades; (b) South and West façades
To avoid damaging the infill wall, all the ACC were bolted to a wooden surface and then
glued to surface of the infill wall, see Figure 5.3 (a). The global behaviour of the RC structure
was captured using: (i) two Piezotronics ACC orthogonally placed in the Northeast and
Southwest corners of each RC slab, Figure 5.2 and Figure 5.3 (b); (ii) motion detecting
cameras placed in the Northwest and Southeast corners of the RC slab of the first storey and
Northeast and Southwest corners of the RC slab of the roof, see Figure 5.4 (a) and (b). Four
motion detection cameras were used, Hamamatsu Photonics C5949 [12], see Figure 5.4 (c),
capable of determining the position of an infrared led each, see Figure 5.4 (d). Both
components, camera and infrared led, are connected to a controller that conditions the
acquired electrical signal, see Figure 5.4 (e). The camera only detects the planar movement of
the infrared led, and each infrared led was connected to the above mentioned corners of the
slabs through steel lever supports, see Figure 5.4 (c).
(a)
(b)
Figure 5.3: PCB Piezotronics accelerometers: (a) at the infill walls; (b) at the corners of the RC slabs
SERIES 227887 MASONRY ENCLOSURES Project
66
It may seem redundant to use ACC and motion detection cameras at the corners of the tested
models since the ACC can provide both accelerations and displacements, but the acquired
data is used for different purposes due to the different sensitivity and resolution of the sensors.
As an example, the ACC are used during the dynamic identifications to obtain dynamic
properties of the model and their evolution during the test, while the motion detection
cameras define the inter-storey drifts more accurately during the seismic tests.
Two extra piezotronics ACC were placed in the RC foundation ring beam, one in each main
direction, in order to compare the acquisitions recorded by the ACC of pre-installed in the
shaking table and the base of the model. These two sets of recordings have to be the same,
excluding possible differences due to intrinsic characteristics of the acquisition equipment,
since the foundation of the model cannot have relative displacements with respect to the
shaking table.
(a)
(b)
(c)
(d)
(e)
Figure 5.4: Hamamatsu photonics c5949 [12]: (a) position of the Hamamatsu leds in the first storey; (b) position of the Hamamatsu leds in the roof; (c) camera and led at the corner of the structure; (d) infrared
led; (e) controller
North
Eas
t
West
South
Hamamatsuposition
North
Eas
t
West
South
Hamamatsuposition
SERIES 227887 MASONRY ENCLOSURES Project
67
The acquisition of the above mentioned equipment (model 1: 44 ACC on the model, 2 ACC
and 2 LVDT on the shaking table and 4 motion displacement cameras; models 2 and 3: 40
ACC on the model, 2 ACC and 2 LVDT on the shaking table and 4 motion displacement
cameras) was carried out using one SCXI chassis connected to a PXI-1001, both from
National Instruments. The ACC channels were conditioned using SCXI-1500 series from
National Instruments and a 481A02 module from PCB Piezotronics, see Figure 5.5 (a). The
acquisition equipment is located in the control room of the laboratory, along with the control
system of the shaking table, see Figure 5.5 (b).
(a)
(b)
Figure 5.5: Acquisition and control room: (a) from top to bottom: NI-SCXI-1001, PCB Piezotronics 481A02 and NI PXI-1052; (b) control room with the shaking table’s controls and the model’s acquisition
system
The K600 Krypton camera [19], see Figure 5.6 (a) to (c), was used on the upper infill wall at
the North façade, see Figure 5.6 (d). It is composed by three CCD cameras that are able to
track the 3D displacement of a set of infrared LED’s, which in this case were placed
throughout the infill wall. The objective was to obtain the out-of-plane deformed shape of the
infill wall, hence two LED’s were placed in the RC beam while the other twenty-two LED’s
were placed on the infill wall. The initial geometrical plan of the LED’s has to be defined,
using the Space Probe, while the acquisition is done using the manufacturers own software, at
an acquisition frequency of 200Hz.
SERIES 227887 MASONRY ENCLOSURES Project
68
(a)
(b)
(c)
(d)
Figure 5.6: K600 Krypton camera: (a) three CCD cameras; (b) Space Probe used to calibrate the initial geometrical plan of the LED’s; (c) acquisition control; (d) distribution of the LED’s along the infill wall on
the upper floor of the North façade
SERIES 227887 MASONRY ENCLOSURES Project
69
5.2 WALL PANELS TEST SETUP
The testing setup for the wall panels simultaneously uses the shaking table, one reaction wall
and the specially designed Testing device for Innovative Masonry infills (TIM), as
represented in Figure 1.4. Such unique testing setup was specifically designed for this test and
is mainly composed of a stiff steel caisson three-dimensional frame, shown in Figure 5.7 and
Figure 5.8, which moves rigidly with the shaking table. It is fixed to the upper beam in the
transverse direction, while a system of rollers allows for an independent motion in the
longitudinal direction (Figure 5.9).
Figure 5.7: Main steel caisson frames of TIM (construction phase)
Figure 5.8: Base columns of main steel frame with detail of bolted connection to the shake table
(construction phase)
SERIES 227887 MASONRY ENCLOSURES Project
70
Figure 5.9: Guiding system of RC frame upper beam and rollers for longitudinal motion (construction
phase)
Both the in-plane and out-of-plane motions should match a given floor response spectra, of
narrow band frequency content. The in-plane motion will enforce an inter-storey drift time-
history on the frame by restraining the upper beam, which is prestressed for withstanding
push-pull actions, and by imposing the displacement of the shaking table on the lower beam.
The upper beam is prevented from moving in the longitudinal direction through a strut
connection to the reaction wall. This connection between the strut and the reaction wall is
performed by a pyramidal support, as depicted in Figure 5.10, which distributes the strut
reaction on the wall. A long rod links the pyramidal support to the RC frame upper beam
through hinged connectors.
All beam-column joints are free to rotate in the plane of the infill, through special hinged base
supports (Figure 5.11 and Figure 5.12). On the other hand, the out-of-plane motion will
consist on a rigid-body vibration of both the upper and lower beams, reproducing the storey
absolute accelerations and thus inducing high-frequency inertia forces perpendicular to the
masonry panel and leading to a local vibration of the infill wall. Note that this shaking table
motion is transmitted to the top beam through the rigid steel caisson. The assembly process of
the test setup is shown in Figure 5.13 to Figure 5.15.
The design of TIM was controlled by the requirement of a very stiff behaviour in the
transverse direction, which was ensured by a vibration frequency above 20 Hz. A parametric
study was conducted using a finite element model with and without the wall panel, as
represented in Figure 5.16, resulting in the modes and frequencies of vibration shown in
Figure 5.17 and Figure 5.18.
SERIES 227887 MASONRY ENCLOSURES Project
71
Figure 5.10: Pyramidal support for strut connection between the RC frame and the reaction wall
Figure 5.11: Hinged base supports for the RC frame specimens (construction phase)
Figure 5.12: Hinged base and pyramidal supports on their final position
SERIES 227887 MASONRY ENCLOSURES Project
72
Figure 5.13: Assembly of TIM components on the shaking table
Figure 5.14: Positioning of TIM over the wall panel setup
Figure 5.15: Complete setup for wall panels tests
SERIES 227887 MASONRY ENCLOSURES Project
73
Figure 5.16: Schematic representation of the finite element models used for the design of TIM, taking into
account (right) or not (left) the wall panels
a) b)
Figure 5.17: Vibration modes of TIM without the wall panel contribution: a) longitudinal (f = 19.9 Hz); b) transverse (f = 33.8 Hz)
a) b) c)
Figure 5.18: Vibration modes of TIM with the wall panel contribution: a) longitudinal (f = 18.4 Hz); b) transverse (f = 23.1 Hz); c) torsional (f = 25.5 Hz)
SERIES 227887 MASONRY ENCLOSURES Project
74
The instrumentation of the wall panels comprises several different sensors:
i) Krypton K600 camera for measuring the out-of-plane deformations of the wall panel
using 16 leds (Figure 5.19)
ii) Video camera for measuring the RC node deformation using data image correlation
methods (Figure 5.20)
iii) Hamamatsu optical system for measuring the horizontal translation of the bottom and
top corners of the RC frame (Figure 5.21)
iv) 36 accelerometers for vibration monitoring of the shaking table, TIM, RC frame and
masonry wall panel (Figure 5.22)
v) Load cells for measuring the dynamic reaction on the strut connecting the reaction
wall and the top beam of the RC frame (Figure 5.23)
Figure 5.19: Out-of-plane wall panel deformation monitoring with Krypton K600 camera
Figure 5.20: Video camera and target points for data image correlation measurement of in-plane
deformations at one RC frame node
SERIES 227887 MASONRY ENCLOSURES Project
75
Figure 5.21: Hamamatsu setup for measuring the horizontal translations of the RC frame nodes
Figure 5.22: Accelerometer setup for RC frame out-of-plane vibration measurements
Figure 5.23: Load cells for strut reaction measurement
SERIES 227887 MASONRY ENCLOSURES Project
76
6 Seismic testing protocol
6.1 TESTING PROCEDURE
The test procedure consisted in a combination of dynamic and seismic tests, the former to
identify the dynamic properties of the buildings and the latter to assess their seismic
performance.
6.2 SHAKING TABLE TUNING
Before conducting a seismic test the shaking table has to be tuned. The tuning procedure is
carried out in an iterative adaptive process starting with a low level intensity and progressing
in several iterations to the final required signals. The procedure follows a sequence of steps:
Dynamic identification of the whole system (ST+model) using a low amplitude “pink”
noise in acceleration to obtain a frequency response function (FRF);
Start with a low amplitude displacement signal input;
Measure displacements and accelerations as output;
Obtain a feedback synthesized displacement signal using a selected cross-over
frequency between measured displacements and accelerations;
Deconvolution of the feedback signal through the system FRF;
Obtain the new “error” signal to use as input signal to the shaking table;
Iterate until input signal is tuned for shaking table;
Increase the input signal and repeat the process for the next stage.
Figure 6.1 shows an example of the definition of the parameters to compute the system
(ST+model) FRF for the shaking table tuning process for the hands on application.
SERIES 227887 MASONRY ENCLOSURES Project
77
Figure 6.1: Shaking table tuning application: definition of parameters
Figure 6.2 shows an example of the system (ST+model) FRF obtained with the “pink” noise
excitation. An example of the signal tuning process is shown in Figure 6.3.
Figure 6.2: Shaking table tuning application: FRF obtained
Input signal (displacement)
Output signals (displacement
and acceleration)
ST-model interaction
Oil column resonance
Delay between input and output
SERIES 227887 MASONRY ENCLOSURES Project
78
Figure 6.3: Signal tuning iterative process
The previous sequence of steps is performed when the model is already placed on the shaking
table. However, the scale and destructive nature of a shaking table experimental program
implies the non-repeatability of any test stage. Therefore, shaking table tests are usually pre-
calibrated with masses representing the model to be tested, see Figure 6.4. This initial
calibration is performed in order to define an adequate input based on the expected behaviour
of the ST+model system. The main issue of using these masses is that they should represent
as closely as possible the inertia distribution of the model.
Figure 6.4: Calibration of the input signals with masses attached to the shaking table
Previous drive
Next drive
Target signal
Acquired signal
Error correction factors
Error signal
Inverse FRF deconvolution
SERIES 227887 MASONRY ENCLOSURES Project
79
6.3 SEISMIC TEST SEQUENCE
The test sequence is obtained applying different scaling factors to the selected accelerograms
(section 7.1), while performing dynamic identification tests in between the different stages.
Table 6.1 describes the seismic test sequence adopted.
Table 6.1 - Shaking table test procedure for the building model
Stage Identification Description
1
DI 0 Initial dynamic identification test
DL Seismic test based on Damage Limitation Limit State
DI1 Dynamic identification test after first stage
2 SD Seismic test based on Significant Damage Limit State
DI2 Dynamic identification test after second stage
3 NC Seismic test based on Near Collapse Limit State
DI 3 Dynamic identification test after third stage
4 1.5xNC
Seismic test with an amplitude of 1.5 times the Near Collapse Limit State stage
DI 4 Dynamic identification test after fourth stage
SERIES 227887 MASONRY ENCLOSURES Project
80
7 Signal generation procedure for the shaking table
tests
7.1 BUILDING MODEL
Shaking table tests can be performed by inputting an earthquake record of a past event,
usually scaled, or an artificial accelerogram. Given the unique and randomness character of a
seismic event, it is difficult to find a suitable earthquake record as far as duration,
accelerations and frequency content are concerned. Using the stochastic tools available in
LNEC-SPA [26], it is possible to generate artificial accelerograms adapted to the code
spectra. In the present experimental work, six artificial accelerograms were generated, whilst
two other were scaled based on the generated ones, in order to obtain four stages of the
shaking table tests, with increasing amplitude, see Table 6.1. The accelerograms of the first
three stages were adapted to the response spectra (damping ratio equal to 5%) of each damage
state described in section 3 of EC8 [30], see Figure 7.1: Damage Limitation (DL - 225 YRP);
Significant Damage (SD - 475 YRP); Near Collapse (NC - 2475 YRP). Here, YRP is the
return period in years. The response spectra for each damage state is obtained by multiplying
the accelerations of the elastic response spectra, which corresponds to the SD state, by the
factor described in EC8 [29].
The last stage was defined as the maximum capacity of the table in terms of velocity, given
the size and mass of the model, and its YRP computed assuming 1.5 and a reference
YRP of 2475. As described in Chapter 3, the maximum weight supported by the table is 392
KN in order to achieve the maximum velocity and acceleration. The weight of the tested
model in the present experimental work (model, foundation RC ring beam and additional
masses) was nearly 434 KN. But this excess of mass did not influence the capacity of the
shaking table to fulfil the input of any of the stages.
SERIES 227887 MASONRY ENCLOSURES Project
81
Transverse direction (East-West)
Longitudinal direction (North-South)
Figure 7.1: Comparison between pseudo-acceleration response spectra of the accelerograms generated and the response spectra, already scaled following the similitude law of Cauchy-Froude, obtained from
EC8 [30]
Given the geographical situation of continental Portugal, in any design situation one obtains
two response spectra as seen in Figure 2.33: (i) type one corresponding to a scenario of a far-
field seismic action; (ii) type two corresponding to a scenario of a near-field seismic action. In
design, the envelope of the response using both spectra is used in order to obtain the seismic
design internal forces. Here, the input signals generated for each stage were adapted only to
type one (far-field seismic action) response spectrum since it provides higher accelerations in
the expected natural frequencies of the models.
One signal was input in each horizontal direction (N-S or longitudinal and E-W or transverse,
see Figure 3.2). The signals were uncorrelated, with approximately the same PGA (peak
ground acceleration), PGV (peak ground velocity) and PGD (peak ground displacement) and
duration of around 30 seconds in the intense phase, see Figure 7.2. Given the Cauchy-Froude
similitude law, see Table 2.2, the acceleration was not scaled ( 1.5 / for seismic
area 1.3) and the frequency, or time, was. Therefore, the generated accelerograms were
adapted to a response spectra with the time reduced by 1.5 .
0.01 0.1 10
2
4
6
8
10
12
14
16
18
20 Generated
input EC8
Acc
eler
atio
n (m
/s2 )
Period (s)
DL 225yrp
SD 475 yrp
NC 2475yrp
1.5xNC 4574yrp
0.01 0.1 10
2
4
6
8
10
12
14
16
18
20 Generated
input EC8
Acc
eler
atio
n (m
/s2 )
Period (s)
DL 225yrp
SD 475 yrp
NC 2475yrp
1.5xNC 4574yrp
SERIES 227887 MASONRY ENCLOSURES Project
82
Transverse direction (East-West)
Longitudinal direction (North-South)
Figure 7.2: Time histories of the input signal of stage 2 (SD 475 YRP) reduced at 1:1.5 scale using Cauchy-Froude’s similitude law (see Table 2.2)
0 5 10 15 20 25 30 35-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Acc
eler
etio
n (m
/s2 )
Time (s)
PGA=1.82 m/s2
0 5 10 15 20 25 30 35-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Acc
eler
atio
n (m
/s2 )
Time (s)
PGA=1.68 m/s2
0 5 10 15 20 25 30 35-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
Vel
ocity
(m
/s)
Time (s)
PGV=0.13 m/s
0 5 10 15 20 25 30 35-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
Vel
ocity
(m
/s)
Time (s)
PGV=0.11 m/s
0 5 10 15 20 25 30 35-25
-20
-15
-10
-5
0
5
10
15
20
25
Dis
plac
emen
t (m
m)
Time (s)
PGD=22.09 mm
0 5 10 15 20 25 30 35-30-25-20-15-10-505
1015202530
Dis
plac
emen
t (m
m)
Time (s)
PGD=26.73 mm
SERIES 227887 MASONRY ENCLOSURES Project
83
7.2 WALL PANELS
The same seismic inputs were used for the wall panels tests as in the building model, except
that these were at full scale. The comparison of the response spectra of both components with
the code one corresponding to the Significant Damage Limit State is represented in
Figure 7.3. These accelerograms were then applied at the base of a representative building
model, see Figure 7.4, in order to obtain the time-history of the wall panel at the 4th floor. The
most important vibration modes for this building are represented in Figure 7.5 for the
longitudinal direction and in Figure 7.6 for the transverse direction.
Figure 7.3: Comparison between pseudo-acceleration response spectra of the accelerograms generated
and the response spectra obtained from EC8 [30]
Figure 7.4: Representative building model for wall panel input time-history definition
SERIES 227887 MASONRY ENCLOSURES Project
84
Figure 7.5: Longitudinal modes of vibration 1 (1.35Hz) and 2 (4.28Hz)
Figure 7.6: Transverse modes of vibration 1 (2.78Hz) and 2 (10.94Hz)
The response time-history in terms of interstorey drift at the 4th floor for the seismic input
considered is shown in Figure 7.7 and corresponds to the shaking table motion to be applied
in the longitudinal direction. On the other hand, the absolute acceleration observed in the out-
of-plane direction is represented in Figure 7.8 and corresponds to the shaking table motion
transmitted by TIM to the masonry panel inside the RC frame.
SERIES 227887 MASONRY ENCLOSURES Project
85
Figure 7.7: Interstorey drift time-history for in-plane motion
Figure 7.8: Absolute acceleration time-history for out-of-plane motion
SERIES 227887 MASONRY ENCLOSURES Project
86
8 Identification technique
The dynamic parameters of the structure (modal frequencies, modal damping and modal
configurations) in the seismic tests are usually obtained by the modal analysis. For this
purpose white noise time-histories with low amplitude or impulse signals are required.
8.1 WHITE NOISE
Before the first stage and after each test stage, the building model is subjected to two inputs,
again orthogonally horizontal and uncorrelated, specially generated with the purpose of
obtaining the dynamic properties of the model (natural frequencies, mode shapes and critical
damping ratios) and their evolution along the experimental test, see Figure 8.1. As these
properties are directly related to the stiffness, the damage state of the structure can be
characterized. Dynamic identification tests under these conditions are normally referred to as
forced vibration tests, in opposition to the usual ambient vibration tests. When compared to
the seismic tests, see Figure 7.2, these accelerograms have lower accelerations and higher
frequency range and duration, and are not adapted to a particular response spectrum but
generated as a low amplitude white noise. Obviously, the dynamic identifications should not
introduce additional damage to the structure, and the maximum amplitude remains relatively
low. There is also a difference in the amplitude of the transverse and longitudinal signals,
with a PGA of 0.44 m/s2 and 0.80 m/s2 respectively, due to the different stiffness of the RC
structure in each direction, see Figure 2.32, which has single bays in the transverse direction
and double bays in the longitudinal direction. Hence, the longitudinal direction has a higher
stiffness and requires larger amplitudes of the input signal.
SERIES 227887 MASONRY ENCLOSURES Project
87
Transverse direction (East-West)
Longitudinal direction (North-South)
Figure 8.1: White noise signals for dynamic identification tests
8.2 IMPULSE SIGNAL
On the other hand, the wall panels’ dynamic identification tests were carried out using
impulsive signals instead. The frequency response gain H fk and phase f k spectra, as
well as the coherence function 2 f k were calculated taking into account the following
approach.
Sequences of data sampled in a time resolution h, are divided into q frames of n points with
the total duration of the sequence given by n h q 1 is then applied.
With the FFT procedure, estimates of one-sided power spectral density functions (psdf) for
the input, output and cross-spectral can be evaluated using the following expressions:
~, ,G
h
NXk i
xk i
2 2 for the input psdf
~, ,G
h
NYk i
yk i
2 2 for the output psdf
~,,
,,G
h
NX Yk i
x yk ik i
2
for the cross psdf between input and output
where X k i,
2 and Yk i,
2 is the squared modulus of FFT components, at frequency f k , for the
input and output respectively and for frame i. The complex conjugate of X k i, is expressed by
X k i, .
0 20 40 60 80 100 120 140 160 180-0.5
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
Acc
eler
atio
n (m
/s2 )
Time (s)
0 20 40 60 80 100 120 140 160 180-0.6-0.5-0.4-0.3-0.2-0.10.00.10.20.30.40.50.6
Acc
eler
atio
n (m
/s2 )
Time (s)
SERIES 227887 MASONRY ENCLOSURES Project
88
Taking into account the averaging process over all the q frames, the expected power spectra
estimates, at frequency f k , for the total duration of the sequences are evaluated by:
~ ~ ~
~ ~ ~
~ ~ ~
, , ,
, , ,
, , ,
, , , ,
Gq
G G G
Gq
G G G
Gq
G G G
kx x x x
y y y y
kx y x y x y x y
k k k q
k k k k q
k k k q
1
1
1
1 2
1 2
1 2
The gain factor and the phase factor of the FRF between input x and output y can now be
carried out using:
H fG
Gk
kx y
kx
,
fG
Gk
kx y
kx y
tanIm
Re
,
,
1
Finally, coherence estimates are evaluated by:
2
2
fG
G Gk
kx y
kx
ky
,
SERIES 227887 MASONRY ENCLOSURES Project
89
9 Analysis of results
9.1 BUILDING MODEL TEST RESULTS
9.1.1 Initial test results
The building model (hereafter termed model 3) is equal to model 2, with the difference that
masonry walls have no bed joint reinforcement but have a reinforced rendering nailed to the
RC frame on both sides. Hence, model 3 represents another possibly future solution for RC
frames with masonry infills. Figure 9.1 presents the position, and label, of the accelerometers
in model 3. In this model, again the accelerometers were placed on the exterior only since the
infill walls have only one leaf.
North South East West
Figure 9.1: Position and label of the accelerometers in model 3
Overall damage and crack patterns
Model 3 was subjected to the four test stages as models 1 and 2, but the fourth was not
successful due to technical problems. The transportation, done using a crane, and the first
testing stage (225 YRP) did not visually damage the mode. After stage 2 (475 YRP) the
model presented cracks in the mortar rendering in all four corners, starting at the base of the
RC column, and between the jambs on the intermediate columns of the East and West
façades, see Figure 9.2. Small cracks starting at the corners and moving towards the RC frame
BNE 2L
BNE 1L
N1 4
N1 1N1 3
N1 2
INP L
N1 5
BSW 2L
S2 6
S2 9
S2 3
S2 7
S2 1S2 4
BSW 1L
S1 4
S1 7
S1 1S1 6S1 3
S1 9
S2 2
S2 8
S2 5
S1 8
S1 2
S1 5E1.1 2
E1.1 3
E1.2 1
E2.1 3
E1.1 1
E2.1 1
E2.1 2
BNE 2T
BNE 1T
E2.2 2E2.2 1
INP T
BSW 2T
BSW 1T
W1.2 1W1.1 1
SERIES 227887 MASONRY ENCLOSURES Project
90
of some of the openings at the ground floor were also visible, while the first floor presented
no visual damage.
After stage 3 (2475 YRP), see Figure 9.3, the cracks in the mortar at the corners of the models
extended and small pieces of mortar rendering fell, see Figure 9.4 (c) and (d). The cracks in
the jambs of the East and West façade also were further extended. New cracks surrounding
the ground floor infills of the North and South façades appeared, along with some cracks in
the infill wall at the first storey of the North façade, mainly between the openings. Overall,
the model presented light damage, see Figure 9.4 (a) and (b), and the cracks, mainly at the
corners, seemed to affect only the mortar.
North South
East West
Figure 9.2: Crack patterns of model 3 after stage 2 (475 YRP) (Note: the drawn lines on the RC frame represent damage on the mortar rendering applied to the RC frame)
SERIES 227887 MASONRY ENCLOSURES Project
91
North South
East West
Figure 9.3: Crack patterns of model 3 after stage 3 (2475 YRP) (Note: the drawn lines on the RC frame represent damage on the mortar rendering applied to the RC frame)
(a) (b)
(c) (d)
Figure 9.4: Damage in model 3 after stage 3 (2475 YRP): (a) infill wall at the ground floor of the North façade; (b) infill wall at the upper floor in the East façade; (c) damaged mortar rendering at the Southeast
corner; (d) damaged mortar rendering at the Southwest corner
SERIES 227887 MASONRY ENCLOSURES Project
92
Modal frequencies of the RC structure
The dynamic identifications performed during the tests of model 3 presented the same quality
as the previous models, with very high coherences between the input and output signals,
leading to the identification of the same five mode shapes as in the previous models, see
Figure 9.5, although the first transverse and first longitudinal changed positions. Since model
3 shares the same geometry with the previous models, and the same structural materials as
model 2, this change is most likely associated to an undesirable interaction between the model
and the shaking table due to different stress levels applied on the connecting bolts or
geometrical imperfections in the RC foundation ring beam. Another reason can be the
interaction between masonry infill and frame, which depends on workmanship and the actual
execution of each infill. As for the other three modes, (first) torsional and second longitudinal
and transverse, the order is the same as in the previous models. Also, as in the previous
models, the torsional mode shape was not possibly correctly captured, see Figure 9.5, as the
roof RC slab is not rotating, while the first floor RC slab is rotating. In the second transverse
mode the South façade also presented a small longitudinal movement.
The variation of the peaks in the FRF’s was followed along the several dynamic
identifications, see Figure 9.6 (a), confirming the good quality of the results and allowing for
damage detection through the frequency decrease in all five modes along the three test stages,
see Figure 9.6 (b). Until the second stage, the longitudinal direction presented no frequency
decrease, the transverse direction presented an average 5.1% frequency decrease and the
torsional mode presented a 5.5% frequency decrease, in comparison to DI0. After stage 3
(2475 YRP), the average decrease in the longitudinal direction was 15.75% and the average
frequency loss in the transverse direction was 24.0%, which is not in agreement with model 2.
A possible reason for this higher decrease in the transverse direction is the slightly higher
recorded PGA in that direction when compared to longitudinal recorded PGA, see Figure 9.24
and Figure 9.25, which is the opposite situation of the previous tests. Another possible reason
is the influence of the interface between the masonry and the frame in the response, as
addressed before. The torsional mode presented a 31.1% frequency decrease, possibly
associated to a loss of connection between the infill walls and the RC frame since the high
frequency of the torsional mode is very dependent on the infill walls. The dynamic data is in
accordance with the observed slight damage.
SERIES 227887 MASONRY ENCLOSURES Project
93
1st Longitudinal Mode (6.26Hz)
1st Transverse Mode (10.01Hz)
Torsional Mode (28.02Hz)
2nd Longitudinal Mode (32.55Hz)
2nd Transverse Mode (35.63Hz)
Figure 9.5: Mode shapes of the DI 0 of model 3 (initial dynamic identification test)
(a) (b)
Figure 9.6: Frequency change along loading stages: (a) variation of the FRF’s along the test of model 3 at the accelerometer BNE – 1L; (b) evolution of the frequencies along the test of model 3 and their final
variation in respect to DI 0
Table 9.1 presents the experimental estimation of the damping ratios along the several
dynamic identifications. As in the previous models, none of the identified mode shapes had
3 4 5 6 71
2
3
4
5
6.26 Hz6.26 Hz
DI 0 DI 1 DI 2 DI 3
5.49 Hz
Gai
n F
acto
r
Frequency (Hz)
1st Longitudinal
6.26 Hz
DI 0 DI 1 DI 2 DI 30
5
10
15
20
25
30
35
40
27.8 Hz(21.9%)
27.2 Hz(16.3%)
35.6 Hz
32.5 Hz
7.4 Hz(26.0%)5.3 Hz
(15.2%)
19.3 Hz(31.1%)
10.0 Hz
6.3 Hz
Freq
uenc
y (H
z)
Dynamic identification
1st Longitudinal
1st Transversal Torsion
2nd Longitudinal
2nd Transversal
28.0 Hz
SERIES 227887 MASONRY ENCLOSURES Project
94
the expected damping increment along the tests, confirming the difficulties in the
experimental estimation of this parameter. The vulnerability curves presented in Figure 9.7
confirm that the longitudinal direction presented no considerable damage until the second
stage, although cracks between the jambs at the ground level of the East and West façades
were visible. After stage 3 (2475 YRP), the transverse direction presented considerably more
damage when compared to the longitudinal one, which is clearly associated to the different
Energy Input, more than 30% higher. Overall, the results are in agreement with the crack
patterns. It is also noted that the damaged indicator reached a level far lower than the other
models, which reached 1.0 (collapse) for model 1 and 0.8 for model 2. This seems to indicate
that the capacity reserve of this model is still much higher than the model 2.
Table 9.1 - Experimental damping ratios of model 3
1st Longitudinal 1st Transverse Torsion 2nd Longitudinal 2nd Transverse
DI 0 (%) 6.12 4.80 1.93 2.14 0.99
DI 1 (%) 5.55 2.62 3.00 1.79 0.87
DI 2 (%) 4.80 16.75 2.99 2.06 2.63
DI 3 (%) 8.15 5.97 4.87 3.86 2.43
Figure 9.7: Seismic vulnerability curves of model 3 in the transverse and longitudinal directions, using the PGA and Input Energy as input
0 1 2 3 4 5 6 7 8 90.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35 1st Transversal Torsional
2nd Transversal
Dam
age
indi
cato
r d
PGA (m/s2)
0 1 2 3 4 50.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35 1st Transversal Torsional
2nd Transversal
Dam
age
indi
cato
r d
Input Energy (J)
0 1 2 3 4 5 6 7 8 90.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Dam
age
indi
cato
r d
PGA (m/s2)
1st Longitudinal
2nd Longitudinal
0 1 2 3 4 50.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Dam
age
indi
cato
r d
Input Energy (J)
1st Longitudinal
2nd Longitudinal
SERIES 227887 MASONRY ENCLOSURES Project
95
Modal frequencies of the infill walls
Figure 9.8 presents the frequency decrease of the infills on South façade and the infill at the
ground floor of the North façade. The blind infill walls have a higher frequency than the one
with openings and the infill wall at the upper level has a higher frequency than the one at the
ground level. Even though the model presented no damage after the first stage (225 YRP), the
infill walls at the ground floor immediately lost stiffness, presumably due to the loss of
connection to the RC frame or some reparation of the stiff rendering from the RC frame. After
stage 2 (475 YRP) the model presented cracks, mainly at the corners and jambs but some at
the openings, and the infill walls of the ground floor presented a frequency decrease, on
average, of 5%. The results are in agreement with the crack patterns, since most of the cracks
seemed to be only at the mortar rendering. Until this point the infill wall at the upper floor of
the South façade presented no frequency decrease.
After stage 3 (2475 YRP), the South façade presented a higher frequency loss when compared
to North one, which is not in total agreement with the crack patterns as the infill wall at the
ground floor of the North façade presented more cracks. The upper infill wall in South façade
presented no cracks and the ground level one presented cracks mainly at the bottom, while the
cracks on the corners of the façade were on the mortar rendering of the RC columns. Overall,
the infill walls did not present extensive cracking which is in accordance with the loss of
stiffness. Again, the frequency loss in this model did not reach the level of model 1 (25% at
stage 3 and then collapse) and model 2 (up to 65%).
Figure 9.8: Evolution of the frequencies of the infill walls in the North and South façades along the test of
model 3 and their final variation in respect to DI 0
DI0 DI1 DI2 DI3
50
55
60
65
70
61.6 Hz(7.6%)
56.5 Hz(11.8%)
53.2 Hz(5.5%)
66.7 Hz
64.1 Hz
56.3 Hz
P1 North facade P1 South facade P2 South facade
Fre
quen
cy (
Hz)
Dynamic identification
SERIES 227887 MASONRY ENCLOSURES Project
96
Interstorey displacements and drifts
Figure 9.9 presents the interstorey displacements and drifts, in the longitudinal and transverse
directions, and in the three test stages. As expected, the displacements and drifts increased
with the seismic amplitude. The ground level recorded similar displacement and drift values
in all three stages, while the roof level recorded slightly higher displacements in the
longitudinal direction, in agreement with the first mode shape, even though the second mode
presented a higher frequency decrease. Both directions recorded increasingly smaller
differences between the ground and first levels, and on the last stage the transverse direction
recorded a higher drift at the upper level. The results are in agreement with the observed
damage which was evenly distributed through the ground level. Note, again, that the drifts are
much lower than in model 1 (0.5% at stage 3 and then collapse) and model 2 (up to 4%).
Longitudinal direction Transverse direction
Figure 9.9: Interstorey displacements and drifts of model 3
PGA of the infill walls and RC structure
Figure 9.10 presents the recorded PGA in the RC structure at the slab levels and in all infill
walls, for each direction and test stage. The PGA recorded at the RC structure increased with
the seismic amplitude, the roof level recorded higher PGA’s when compared to the first floor
0 2 4 6 8 10 12
storey 1(2 meters)
stage 1 stage 2 stage 3
Displacement (mm)
roof(4 meters)
00 1 2 3 4 5 6 7 8 9 10
stage 1 stage 2 stage 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0.0 0.1 0.2 0.3 0.4
stage 1 stage 2 stage 3
storey 1(2 meters)
Drift (%)
roof(4 meters)
00.0 0.1 0.2 0.3 0.4
stage 1 stage 2 stage 3
0
storey 1(2 meters)
roof(4 meters)
Drift (%)
SERIES 227887 MASONRY ENCLOSURES Project
97
RC slab, and the highest values in the second and third stages (225 and 475 YRP) were
recorded in the longitudinal and transverse directions, respectively. The amplifications in the
RC frame present a moderate change, with the exception of the roof in the longitudinal
direction that almost was halved. For the masonry infills, the longitudinal direction presenting
a slight increase from stage 1 (225 YRP) to stage 2 (475 YRP) and a 42.2% decrease from
stage 2 to stage 3 (2475 YRP), even if not consistently for all walls, while the transverse
direction presented a nearly constant amplification. Again, this confirms the observation in all
models that amplifications do not follow a clear trend, and the initial amplification provides a
reasonable estimate of the dynamic response in the non-linear range.
The PGA recorded at the infill walls also increased with the seismic amplitude. On the
transverse direction, the infill walls at the first storey recorded higher values in all stages but
on the longitudinal direction the maximum PGA’s were recorded in the infill wall at the
ground floor of the North façade. The longitudinal direction recorded higher values when
compared with the transverse direction, even though the input PGA was higher in the
transverse direction, because of the considerably higher amplifications.
Longitudinal direction Transverse direction
RC structure
Figure 9.10: Recorded PGA and amplifications at the infill walls and at the RC structure for each test stage of model 3
7 6 5 4 3 2 1 00 5 10 15 20 25
1
2
3
N1 S1 S2
Amplification Acceleration (m/s2)
Stage
3.0 2.5 2.0 1.5 1.0 0.5 0.00.0 2.5 5.0 7.5 10.0 12.5 15.0
1
2
3
E 1.1 E 1.2 E 2.1 E 2.2 W1.1 W 1.2
Amplification
Stage
Acceleration (m/s2)
5 4 3 2 1 00 5 10 15 20
1
2
3
Storey 1 Trans Dir Storey 1 Long Dir Roof Trans Dir Roof Long Dir
Amplification Acceleration (m/s2)
Stage
SERIES 227887 MASONRY ENCLOSURES Project
98
Out-of-plane PGD and deformation of the infill walls
Figure 9.11 and Figure 9.12 present the out-of-plane deformation of the infill walls of the
South façade in all test stages, with the first storey infill wall presenting a similar the same
deformed shape, with the maximum deformation at the centre and lower part of the wall, and
PGD in the first (225 YRP) and second stages (475 YRP). In the third stage (2475 YRP), the
maximum deformation area was slightly smaller and the PGD remained the same as in the
previous stages. This unaltered state along the test was expected as the infill wall did not
present any cracks and only a 5.6% frequency decrease, in respect to DI0, was recorded in
stage 3. As for the infill wall at the ground floor, in the first and second stages the wall
presented the same deformed shape and similar PGD, with the maximum value recorded at
the top of the wall, but on the third stage the wall presented a higher flexibility, with at least
twice the displacements of the previous stages, and an increment in the PGD area. This infill
wall presented the highest frequency decrease, in the longitudinal direction, and cracks
surrounding the frame, which confirms the agreement between the results.
Figure 9.13 presents the out-of-plane deformation of the infill wall at the upper floor of the
North façade, with the PGD increasing in all stages and with the area of higher displacements
concentrated between the openings. These results are in agreement with the observed damage,
as the cracks in this infill wall were also concentrated between the openings and started in
stage 2. The PGD of the rest of the infill walls of model 3 are presented in Figure 9.14, where
all infills present an increment in the displacement with the seismic amplitude and the infill
walls at the upper level present higher displacements than infill walls at the ground floor.
SERIES 227887 MASONRY ENCLOSURES Project
99
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.11: Out-of-plane deformation of the infill wall at the ground level of the South façade of model 3 in mm
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.12: Out-of-plane deformation of the infill wall at the first storey of the South façade of model 3 in mm
13.3411.44
9.5317.625 5.719
3.8131.906
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
13.2311.35
9.475 7.600 5.725 3.8501.975
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
13.29
13.2911.33
9.3637.4005.438
3.4751.513
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
1.3253.300
5.275 7.2509.22511.2013.18
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
-0.33131.638
3.6065.5757.5449.513
11.48
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
24.98
26.34
23.61
27.7029.0630.4331.79
26.34
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
SERIES 227887 MASONRY ENCLOSURES Project
100
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.13: Out-of-plane deformation of the infill wall at the first storey of the North façade of model 3 in mm.
Figure 9.14: Out-of-plane PGD of the North, East and West infill walls of model 3 in mm
9.1.2 Complementary test results
As stated before, due to technical problems in the shaking table, model 3 was only tested until
stage 3 (2475 YRP) but given the light overall damage presented by the structure and infill
walls, both in absolute value and when compared with the other models, the model was
submitted to the first three stages again. The model was not removed from the shaking table,
as the retest was performed the day after the first test, so no damage was introduced in the
transportation and no changes were made to the boundary conditions. The results of these new
three stages, in the model hereafter denominated as model 3B, cannot be directly compared to
0.8540
0.76650.85400.6790
0.9415
0.5915
1.0291.1170.7665
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
0.300.50
0.69
0.89
1.11.3 1.5
0.50
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
1.11.5
1.9
2.2
2.63.0
3.4
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
0 5 10 15 20 25 30 35
1
2
3
E 1.1 E 1.2 E 2.1 E 2.2 W 1.1 W 1.2 N 1
Sta
ge
Out-of-plane PGD (mm)
SERIES 227887 MASONRY ENCLOSURES Project
101
the previous three shaking table tests as far as the values of the parameters are concerned, but
the damage pattern of the RC structure and infill walls and the collapse mode developed after
the structure has been severely damaged can be compared.
Overall damage and rendering removal
After the three test stages, model 3B still did not present as much damage as the previous
models, see Figure 9.15 (a) and (b), as no new cracks appeared but the ones at the corners, see
Figure 9.15 (c) and (d), and jambs, see Figure 9.15 (e) and (f), only at the ground floor,
widened considerably and parts of mortar rendering were expelled. The reinforced plaster
became loose, as if completely disconnected from the infill walls, and as if it was standing
only because of the connection provided by the additional masses, which is unreasonable and
not a true structural feature. This also confirms the importance of the connections of
reinforced plaster to the walls, which should cross the wall, and not only use nails, as done
here. The upper level presented no significant damage with only small cracks at the corners of
the openings.
SERIES 227887 MASONRY ENCLOSURES Project
102
(a) (b)
(c) (d)
(e) (f) Figure 9.15: Damage in model 3B after stage 3 (2475 YRP): (a) North façade; (b) South façade; (c) crack
and mortar rendering loss at the Northwest corner; (d) crack and mortar rendering loss at Northeast corner; (e) crack at the a lateral jamb in the infill wall at the ground floor of the East façade; (f) crack at
the interior jambs in the infill wall at the ground floor of the East façade
After testing, the additional masses bolted to the infill walls were removed, and as these were
working simultaneously as an attachment of the reinforced plaster to the infill wall, it was
possible to simply remove the reinforced rendering as a whole on both sides of the infill
walls, without the use of any specific equipment. This confirmed that the reinforced plaster
was completely detached from the infill walls and that the fixings of the additional masses
worked as connectors, preventing the rendering from collapse. Moreover, careful analysis of
the un-plastered infill walls showed that these presented limited damage, see Figure 9.16 (a)
and (b), but were mostly disconnected from the RC frame, see Figure 9.16 (c) and (d). Hence,
there was a major contribution reinforced plaster was preventing the out-of-plane collapse of
the infill walls. As for the RC structure, no cracks were detected at mid-height of the RC
SERIES 227887 MASONRY ENCLOSURES Project
103
columns, but only at the upper connection to the beams, see Figure 9.16 (e). The RC was very
flexible under these conditions, meaning that the reinforced plaster was also rather important
in preventing the collapse of the entire system.
(a) (b)
(c) (d)
(e) Figure 9.16: Damage in the infill walls and RC structure after the reinforced rendering removal at the ground floor: (a) infill wall of the North façade; (b) South infill wall with a compression crush at right
down corner; (c) gap between one of the West the infill wall and RC frame in the West wall; (d) infill walls of the West façade; (e) extensive cracking at the upper column-beam connection in the Northwest corner
Modal frequencies of the RC structure and infill walls
Figure 9.17 presents the frequency decrease of the RC structure and of three infill walls in
model 3B along the three test stages. The initial dynamic identification, DI0, corresponds to
DI3 in model 3, and the percentage at the last frequency is the total loss since DI0 of model 3,
SERIES 227887 MASONRY ENCLOSURES Project
104
which is the undamaged state of the model. In the RC structure, the identification of the (first)
torsional and second longitudinal and transverse modes was not possible after stage 3 (2475
YRP). The first order transverse mode presented a higher stiffness loss when compared to the
longitudinal mode. After stage 2 (475 YRP) the modes presented 0.2Hz of difference between
them and after stage 3 the transverse direction presented a higher stiffness loss. At the end of
the test, the transverse direction presented a 73.5% frequency loss, while the longitudinal
direction presented a 48.6% decrease, when compared to the undamaged state. This difference
can be associated to the infill walls at the ground floor, which on the transverse direction
presented a clear disconnection to the RC frame at the jambs, while on the transverse
direction the infill walls also presented corner crushing due to compression and extensive
damage around the door.
The infill wall at the upper level of the South façade presented an extra 12.2% of frequency
decrease during the three stages, which add to the previous damage totalizing 19.8%, when
compared to the undamaged state. This infill wall did not present any visible damage, and the
loss of stiffness is associated to the loss of connection between the infill wall and the RC
frame. The infill walls at the ground floor presented a similar and considerably higher
frequency loss at the end of stage 3, when compared to the upper level one, although the
South infill wall presented the highest loss at the first stage (225 YRP) and the North one
presented the highest loss in the last stage. The stiffness loss presented is in agreement with
the observed crack pattern after the reinforced plaster removal.
Figure 9.17: Evolution of the frequencies along the test of model 3B, and their final variation in respect to DI 0 of model 3, at the RC structure and infill walls in South façade and ground level of the North façade
DI 0 DI 1 DI 2 DI 30.02.55.07.5
10.012.515.017.520.022.525.027.530.0
27.5 Hz(22.7%)
26.3 Hz(19.3%)
27.8 Hz27.2 Hz
3.2 Hz(48.6%)
2.7 Hz(73.5%)
17.6 Hz(37.3%)
7.4 Hz
5.5 Hz
Freq
uenc
y (H
z)
Dynamic identification
1st Longitudinal
1st Transversal Torsion
2nd Longitudinal
2nd Transversal19.3 Hz
DI0 DI1 DI2 DI320
25
30
35
40
45
50
55
60
65
53.4 Hz(19.8%)
26.1 Hz(53.7%)
22.1 Hz(65.6%)
61.6 Hz
56.5 Hz
53.2 Hz
P1 North facade P1 South facade P2 South facade
Fre
quen
cy (
Hz)
Dynamic identification
SERIES 227887 MASONRY ENCLOSURES Project
105
Interstorey displacements and drifts
The interstorey displacements and drifts increased with the seismic amplitude, see
Figure 9.18, until the second stage (475 YRP) the longitudinal and transverse directions
presented similar values and the ground level values had a tendency to increase when
compared to the upper level ones. On stage 3 (2475 YRP) the ground level recorded higher
displacements and drifts than the first level. While the first level values were similar in both
direction, the ground floor of the transverse direction presented higher displacements and
drifts. These results are in agreement with the observed crack patterns and dynamic data, as
the transverse direction presented the highest stiffness loss and the corner crushing due to
compression at the ground infill wall of the South façade is associated to in-plane movement
of the RC frame, in this case in the transverse direction. Again, note that the maximum drift at
this stage is in the range of 2%.
Longitudinal direction Transverse direction
Figure 9.18: Interstorey displacements and drifts of model 3B
0 2 4 6 8 10 12 14 16 18 20 22
storey 1(2 meters)
stage 1 stage 2 stage 3
Displacement (mm)
roof(4 meters)
00 5 10 15 20 25 30 35 40
stage 1 stage 2 stage 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0.0 0.5 1.0 1.5 2.0
stage 1 stage 2 stage 3
storey 1(2 meters)
Drift (%)
roof(4 meters)
00.0 0.5 1.0 1.5 2.0
stage 1 stage 2 stage 3
0
storey 1(2 meters)
roof(4 meters)
Drift (%)
SERIES 227887 MASONRY ENCLOSURES Project
106
PGA of the infill walls and RC structure
The PGA values recorded on the RC structure and infill walls, see Figure 9.19, presented no
surprises since there was an increment with the seismic amplitude and the upper levels of the
infill walls recorded higher values when compared to the ground floor ones. As for the
amplifications, there is always a small increment in the recorded values from stage 1 (225
YRP) to stage 2 (475 YRP) and a larger decrement from stage 2 to stage 3 (2475 YRP). In the
RC structure, the transverse direction did not present a higher amplification loss due to the
stiffness loss. The infills at the first level of the transverse direction presented the highest
amplification loss, while on the longitudinal direction it were the infills on the South façade
that presented the highest amplification loss.
Longitudinal direction Transverse direction
RC structure
Figure 9.19: Recorded PGA and amplifications at the infill walls and at the RC structure for each test stage of model 3B
Out-of plane PGD and deformation of the infill walls
Contrary to what happened during the test of model 3, in model 3B the damage presented by
the infill walls along the 3 test stages altered their out-of-plane deformed shape. The infill
wall at the ground level of the South façade, see Figure 9.20, presented very similar values in
7 6 5 4 3 2 1 00 5 10 15 20 25
1
2
3
N1 S1 S2
Amplification Acceleration (m/s2)
Stage
4 3 2 1 00 2 4 6 8 10 12 14 16
1
2
3
E 1.1 E 1.2 E 2.1 E 2.2 W1.1 W 1.2
Amplification
Stage
Acceleration (m/s2)
5 4 3 2 1 00 5 10 15 20
1
2
3
Storey 1 Trans Dir Storey 1 Long Dir Roof Trans Dir Roof Long Dir
Amplification Acceleration (m/s2)
Stage
SERIES 227887 MASONRY ENCLOSURES Project
107
the first and second stage (225 and 475 YRP) with the maximum values at the lower part of
the wall in the first stage and then concentrating at lower lateral sides and mostly and the
corner with the crushed corner, see Figure 9.16 (b). During stage 3 (2475 YRP) the
displacements increased diagonally from the lower left corner to the upper right corner. The
infill wall at the upper level of the South façade, see Figure 9.21, which did not present any
visible damage at the end of the test, presented the highest deformation at the top in the first
stage and on stage 2 and 3 at the sides and top. The infill wall at the upper level of the North
façade, see Figure 9.22, still presented the highest deformation between the jambs, as during
the tests of model 3, in the stage 1, and on stage 2 and stage 3 moved to the lower and left
side, respectively, even though the only visible damage on the infill were very small cracks
connecting both openings.
Figure 9.23 presents the PGD recorded at all other infill walls of model 3B, with the infill
walls of the upper level recording the highest value in all stages and the infill wall at the
ground floor of the North façade presenting the highest PGD of the ground floor infill walls,
within the same values of the South infill wall.
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.20: Out-of-plane deformation of the infill wall at the ground level of the South façade of model 3B in mm
9.1007.800
6.500 5.200 3.9002.600 1.300
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
7.4256.163
4.900
8.688
3.6382.3751.113
8.688
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
8.605
9.600
7.610
10.59
6.615
11.5912.58
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
SERIES 227887 MASONRY ENCLOSURES Project
108
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.21: Out-of-plane deformation of the infill wall at the first storey of the South façade of model 3B in mm
Stage 1 (225 YRP) Stage 2 (475 YRP)
Stage 3 (2475 YRP)
Figure 9.22: Out-of-plane deformation of the infill wall at the first storey of the North façade of model 3B
in mm
2.2803.360
4.440
5.5206.6007.6808.760
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
8.980
8.980
9.630
8.330
10.2810.9311.5812.23
8.980
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)31.23
30.51
30.51
31.23
31.9432.66
29.80
33.3734.09
31.94
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
Infill length (m)
Infi
ll h
eigt
h (m
)
2.72.9
3.1
2.7 2.4
2.7
2.2 2.01.7
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
3.1
1.1
1.1
3.1
5.27.39.3
1.1
11 135.2
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
169.28
168.82
168.35
167.89
167.42
166.95166.49
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.50.00
0.25
0.50
0.75
1.00
1.25
1.50
openingopeningopening
Infill length (m)
Infi
ll h
eigh
t (m
) window window
SERIES 227887 MASONRY ENCLOSURES Project
109
Figure 9.23: Out-of-plane PGD of the North, East and West infill walls of model 3B in mm
9.1.3 Comparison of test results
Following, a comparison of the experimental results of all the models is presented but given
that not all were subjected to the last stage or collapsed during it, the comparison is done
considering the first three stages only.
Target/acquired comparison and shaking table performance
The comparison between the target (input), presented in the previous chapter, and the data
acquired (output) by the accelerometers placed in the shaking table, also described in the
previous chapter, was done using the Peak Ground Acceleration (PGA) and a set of five
integral parameters (Root Mean Square Acceleration (RMSA), Arias Intensity (AI) and Input
Energy (IE) [18], [7] in each of the two main horizontal directions (longitudinal or North-
South, and transverse or East-West), as follows:
(5)
1 (6)
2
(7)
0 5 10 15 20 25 30 35 40 45 50
1
2
3
E 1.1 E 1.2 E 2.1 E 2.2 W 1.1 W 1.2 N 1
Sta
ge
Out-of-plane PGD (mm)
SERIES 227887 MASONRY ENCLOSURES Project
110
(8)
Here, is the time history acceleration, is the duration of the signal or earthquake, is
the gravity acceleration, is the mass of the model and is the time history of velocity.
This comparison is needed despite the calibration tests performed with inert masses since the
models are not inert and, therefore, they influence the behaviour of the shaking table. The
parameters used for the comparison were chosen due to their importance in seismic
engineering and structural dynamics. The PGA is commonly used in standards for design
purposes, even if associated to a response spectrum. A single peak value cannot accurately
represent a seismic action, as seismic actions with the same PGA can result in different
damage scenarios, hence the use of integral parameters. All integral computations depend on
the duration of the seismic action, a parameter with more importance regarding the level of
destruction than the amplitude of the accelerations, particularly for unreinforced masonry
structures. A seismic action with shorter duration and higher accelerations (e.g., Ancona, Italy
in 1972) is likely to be less destructive than a seismic action with longer duration and lower
accelerations (e.g., Mexico City, Mexico in 1985).
As far as the PGA is concerned, see Figure 9.24 (a) and Figure 9.25 (a), and analysing all
models simultaneously, the transverse direction of the shaking table outperformed the
longitudinal direction, since smaller differences were recorded with respect to the target. In
the longitudinal direction the acquired PGA of the models was, on average, 42% higher in
respect to the target, and on the transverse direction the acquired PGA was only 20% higher
than the target value. These computations were done considering all four stages for models 1
and 2 and only the first three stages for models 3 and 3B. Model 3 was subjected to the fourth
stage, but as it can be seen in Figure 9.24 (a) and Figure 9.25 (a), the acquired PGA was 62%
lower in respect to the target in the longitudinal direction and 4% higher in the transverse
direction. This was due to a technical problem in the shaking table and therefore this stage
was not considered. Model 3B was not subjected to the fourth stage due its extensive damage
and possible collapse. But technical problems remained since the recorded PGA in the
transverse direction of stage 3 was twice the value of the target, thus larger than the other
models for this direction.
SERIES 227887 MASONRY ENCLOSURES Project
111
If the comparison between the target and the acquired response is done using the RMSA, see
Figure 9.24 (b) and Figure 9.25 (b), the deviations are considerably lower as the RMSA is not
so dependent on peak values. The average deviation of all models was the same for the
longitudinal and transverse directions and equal to 2%. The AI, see Figure 9.24(c) and
Figure 9.25 (c), is a measure of the earthquake destructiveness based on the RMSA but more
dependent on peak values therefore the deviations increase to 34% and 32% in the
longitudinal and transverse directions, respectively. The IE, see Figure 9.24 (d) and
Figure 9.25 (d), is the only parameter in which the deviation is negative, meaning the acquired
response was lower than the target. The average deviation for the models was 10% and 2% in
the longitudinal and transverse directions, respectively. The integral parameters confirmed the
technical problems above referred during the fourth stage of model 3, excluding it from the
present analysis.
In conclusion, the differences between the acquired and target data is within acceptable limits
in parameters dependant on peak values and good considering parameters more dependent on
the duration of the motion. The better results are in the transverse direction, when compared
to the longitudinal one, due to the characteristics of the shaking table, already presented in the
previous chapter, which has two actuators in the transverse direction and only one actuator in
the longitudinal direction. Therefore, the transverse direction is more sensitive and able to
better replicate the target.
SERIES 227887 MASONRY ENCLOSURES Project
112
(a)
(b)
(c)
(d)
Figure 9.24: Longitudinal direction target/acquired comparison (Fourier filter: 1-40Hz): (a) PGA; (b) Root Mean Square Acceleration; (c) Arias Intensity: (d) Input Energy
(a)
(b)
(c)
(d)
Figure 9.25: Transverse direction target/acquired comparison (Fourier filter: 1-40Hz): (a) PGA; (b) Root Mean Square Acceleration; (c) Arias Intensity: (d) Input Energy
Stage 1 Stage 2 Stage 3 Stage 40
2
4
6
8
10
12
Target Model 1 Model 2 Model 3 Model 3B
PGA
(m
/s2 )
EC8 Design
Stage 1 Stage 2 Stage 3 Stage 40
5
10
15
20
25
30
Ari
as I
nten
sity
(m
/s)
Target Model 1 Model 2 Model 3 Model 3B
Stage 1 Stage 2 Stage 3 Stage 40
5
10
15
20
25
30
Ari
as I
nten
sity
(m
/s)
Target Model 1 Model 2 Model 3 Model 3B
Stage 1 Stage 2 Stage 3 Stage 40
2
4
6
8
Inpu
t Ene
rgy
(J)
Target Model 1 Model 2 Model 3 Model 3B
Stage 1 Stage 2 Stage 3 Stage 40
2
4
6
8
10
12
PGA
(m
/s2 )
Target Model 1 Model 2 Model 3A Model 3B
EC8 Design
Satge 1 Stage 2 Stage 3 Stage 40.0
0.5
1.0
1.5
2.0
2.5
Roo
t Mea
n S
quar
e (m
/s2 )
Target Model 1 Model 2 Model 3 Model 3B
Stage 1 Stage 2 Stage 3 Stage 40
5
10
15
20
25
30
35
Ari
as I
nten
sity
(m
/s)
Target Model 1 Model 2 Model 3 Model 3B
Stage 1 Stage 2 Stage 3 Stage 40
2
4
6
8
Inpu
t Ene
rgy
(J)
Target Model 1 Model 2 Model 3 Model 3B
SERIES 227887 MASONRY ENCLOSURES Project
113
Modal frequencies of the structure and infill walls
For the sake of simplicity, and given the amount of data already provided for each model
individually, the comparison of the modal frequencies and their variation is done through the
vulnerability curves that relate the damage indicator, , with the input energy applied at the
base and in the case of the RC structure only for the first mode in each main direction.
Regarding the seismic vulnerability curves of the RC structure, see Figure 9.26, and on the
transverse direction, model 3 presented a higher damage indicator when compared to models
1 and 2 in stage 2, 27% and 49% respectively, and in stage 3, 35% and 49% respectively,
which is unexpected due its higher resistance concrete and rebar and reinforced plaster infills.
On stage 2 all three models were subjected to the same input energy at the base, but on the
third stage model 3 was subjected to an input energy 16% and 14% higher than models 1 and
2, respectively, which can justify the considerably higher damage indicator of model 3. Model
1 presented a damage factor higher than model 2 in all three stages, as expected given the
models’ lower strength materials and weaker unreinforced infills.
In the longitudinal direction, model 3 presents the lowest damage indicator in all three stages,
but on stage 3 the model was subjected to an input energy 28% lower than models 1 and 2,
hence the 64% and 72% lower damage indicator, respectively, due to mechanical issues
already addressed. Model 1 presented a damage indicator 80% higher in stage 2, when
compared to model 2, but 25% lower on stage 3 which is unexpected. Both models were
subjected to stage 4, and while the model 1 collapsed, which means a unitary damage
indicator, model 2 did not collapse even though it was heavily damaged. One of the factors
that contributed to this improved performance of model 2 is the continuous connection of the
reinforced infills to the RC frame provided by the metallic connecters, while on model 1 all
infills collapsed out-of-plane and the RC structure failed shortly after.
Transverse direction Longitudinal direction
Figure 9.26: Vulnerability curves of the 1st mode in each main direction of the RC structure of the three tested models
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.50.00
0.05
0.10
0.15
0.20
0.25
0.30Stage 3 Model 1
Model 2 Model 3
Dam
age
indi
cato
r d
Input Energy (J)
Stage 2
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00.0
0.1
0.2
0.3
0.4
0.5 Model 1 Model 2 Model 3
Dam
age
indi
cato
r d
Input Energy (J)
Stage 3
Stage 2
SERIES 227887 MASONRY ENCLOSURES Project
114
Figure 9.27 presents the average vulnerability curves of the infill walls on the North façade,
with openings, and South façade, without openings, of all three tested models and in the first
three stages. The same issues with the input energies are patent here, with model 3 being
subjected a 28% lower input in stage 3 when compared to models 1 and 2. As far as the North
infills are concerned, on the second stage model 3 presented the highest damage indicator and
models 1 and 2 similar values, while on stage 3 model 3 presented a damage indicator 71%
and 63% lower than models 1 and 3, respectively. From stage 2 to stage 3, the bed joint
reinforced infill walls presented a damage indicator 21% higher when compared to the
unreinforced ones, even though the unreinforced ones collapsed during the last stage and the
reinforced ones did not, meaning that the bed joint reinforcement does not contribute to the
decrease of damage but rather prevent the out-of-plane collapse of the infill. In the South
façade, and in stage 2, the unreinforced infill walls presented a higher damage coefficient
when compared to the reinforced ones, while the bed joint reinforced infills presented a
higher damage indicator than the plaster reinforced ones. This order was kept unaltered from
stage 2 to stage 3, as well as the percentage difference between the models.
North infill walls South infill walls
Figure 9.27: Average vulnerability curves of the North and South infill walls of the three tested models
Interstorey displacements and drifts
Figure 9.28 presents the interstorey displacements of the first three stages of the tested models
in both main directions. In the second stage, corresponding to the design earthquake, and in
the transverse direction, model 1 presented displacements 70.4% and 13.8% higher, on
average, than models 2 and 3, respectively, as expected since it has a lower strength concrete
and rebar and it was designed with the older design standard [36] instead of the Eurocodes
[29] and [30]. On the third stage, model 1 presented a displacement 20.6% and 51.9% lower,
on average, than models 2 and 3, respectively. Model 3 presented displacements 56.6%
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00.00
0.05
0.10
0.15
0.20
Dam
age
indi
cato
r d
Input Energy (J)
Model 1 Model 2 Model 3
Stage 2
Stage 3
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00.00
0.05
0.10
0.15
0.20
Dam
age
indi
cato
r d
Input Energy (J)
Model 1 Model 2 Model 3
Stage 2
Stage 3
SERIES 227887 MASONRY ENCLOSURES Project
115
higher, on average, during stage 2 and 31.3% higher, on average, during stage 3. In the
longitudinal direction, model 1 did not present the highest displacement on stage 2, but it also
presented the lowest displacements on stage 3. As for modes 2 and 3, in the second stage the
difference between both is 36.4%, with model 3 presenting higher displacements, and in the
third stage the difference is 20.6%, with model 2 presenting higher displacements. This
inversion is mostly associated to the lower input energy in model 3 during the third stage,
which led to lower damage in model 3. The use of different design standards between model 1
and models 2 and 3 is very clear in both directions during stage three, with model 1 presenting
lower displacement capacity or ductility than the other two models.
Figure 9.29 presents the interstorey drifts of the first three stages of the tested models in both
main directions. In the second stage all models present a similar drift on both storeys, or the
second floor has a higher drift, while on stage 3 the ground floor tends to have higher drifts,
introducing considerably more in-plane damage in the infill walls and reducing their out-of-
plane capacity. Model 2 presents lower drifts, when compared to model 3, along the two
stages which indicate that the bed joint reinforcement might contribute to higher in-plane
stiffness when compared to the reinforced plaster. The absence of a clear trend of model 1
regarding the other two models, confirms that the strength of the RC structure and the
presence of reinforcement on the infill walls has a low influence on the in-plane behaviour of
the frames, aggravated by the fact that the most influential parameters, which are the
geometry of the frame and the geometry, presence of openings and strength of the infill, are
the same in all three tested models.
SERIES 227887 MASONRY ENCLOSURES Project
116
Transverse direction Longitudinal direction S
tage
2
Sta
ge 3
Figure 9.28: Interstorey displacements of the three tested models in the transverse and longitudinal directions
Transverse direction Longitudinal direction
Sta
ge 2
S
tage
3
Figure 9.29: Interstorey drifts of the three tested models in the transverse and longitudinal directions
0 1 2 3 4 5 6 7
Model 1 Model 2 Model 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
00 1 2 3 4 5
Model 1 Model 2 Model 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0 1 2 3 4 5 6 7 8 9
Model 1 Model 2 Model 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
00 1 2 3 4 5 6 7 8 9 10 11 12
Model 1 Model 2 Model 3
storey 1(2 meters)
Displacement (mm)
roof(4 meters)
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6
roof(4 meters)
Model 1 Model 2 Model 3
0
storey 1(2 meters)
Drift (%)
0.0 0.1 0.2
Model 1 Model 2 Model 3
storey 1(2 meters)
Drift (%)
roof(4 meters)
0
0.0 0.1 0.2 0.3 0.4
roof(4 meters)
Model 1 Model 2 Model 3
0
storey 1(2 meters)
Drift (%)
0.0 0.1 0.2 0.3 0.4
Model 1 Model 2 Model 3
storey 1(2 meters)
Drift (%)
roof(4 meters)
0
SERIES 227887 MASONRY ENCLOSURES Project
117
PGA of the RC structure and infill walls
Figure 9.30 presents the maximum recorded acceleration and amplification regarding the
input acceleration, on average, for the infills walls in each façade and the RC structure in each
main direction for the three tested models. Model 3 presents higher accelerations and
amplifications along the test, when compared to models 1 and 2, which is unexpected and
possibly unrelated to the characteristics of the model but rather associated to the mechanical
problems above mentioned. As far as the infills are concerned, another justification is
associated to the bolting of the accelerometers to the infills, which is actually only bolted to
the reinforced plaster, and as seen in Figure 9.15 and Figure 9.16 detached from the infills and
remained attached to the RC structure. This means that at later stages along the test some
accelerometers are measuring only the accelerations of a stiff piece of reinforced mortar
plaster and not the actual infill wall, leading to higher accelerations and amplifications.
Analysing models 1 and 2, in the longitudinal direction, the infill walls with openings at the
North façade the unreinforced infill walls of model 1 presented higher amplification when
compared to the reinforced ones of model 2, but lower maximum acceleration, while the blind
infill walls of the South façade present similar results. The infill walls with openings in the
East and West façade follow the ones in the North façade, especially in stage 3, with the
unreinforced infills of model 1 presenting higher amplifications and lower maximum
accelerations when compared to the bed joint reinforced infill walls of model 2.
As for the RC structure, and analysing again models 1 and 2, no trend can be found and both
models present similar results, and the justification is associated to the similarities of both
models regarding the most influential parameters.
SERIES 227887 MASONRY ENCLOSURES Project
118
North façade South façade
East façade West façade
RC structure Storey 1 RC structure roof
Figure 9.30: Average recorded PGA and amplifications at the infill walls and at the RC structure for each test stage of all tested models
6 5 4 3 2 1 00 5 10 15 20 25 30
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
5 4 3 2 1 00 5 10 15 20 25
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
2 1 00 2 4 6 8 10 12
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
2 1 00 2 4 6 8 10
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
2 1 00 5 10 15
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
4 3 2 1 00 5 10 15 20
1
2
3
Model 1 Model 2 Model 3
Amplification Acceleration (m/s2)
Stage
SERIES 227887 MASONRY ENCLOSURES Project
119
10 Main conclusions
Model 1 had a good performance during the seismic standard PGA of stage 2 (475 YRP) with
no visible damage, even though the dynamic data presented frequency loss, on the RC
structure and infill walls. The soft-storey collapse mechanism developed during the fourth
stage is highly undesirable, with a low energy dissipation capacity and brittle collapse [40]. It
was not clear if the collapse mechanism developed was due to the detailing of the RC
structure imposed by the design standard [35] or the influence of the infill walls, since RC
columns also presented hinges at mid-height just before the collapse of the structure. Some of
the thin blocs applied in the RC columns and beams to avoid thermal bridges, a very common
solution in the Portuguese built patrimony, cracked and fell during stage 3 (2475 YRP). The
double leaf unreinforced infill walls underperformed during the last stage (4574 YRP),
collapsing out-of-plane by rotating as a rigid body around the base line of the model. The
interior and exterior leaves presented a similar seismic behaviour.
Model 2 presented a good seismic performance when subjected to the seismic standard PGA
during stage 2 (475 YRP), although the RC structure registered a small decrease in the modal
frequencies. The model did not collapse during the last stage (4574 YRP) but presented
severe, and most likely irreparable, damage and the RC structure developed a soft-storey
mechanism. The seismic standard used in the design of model 2 [29], [30] clearly details the
structure in order for the development of a beam-sway mechanism [40] by forcing the hinges
to appear at the beams and not the columns. Attending also the mid-height cracks found in all
the RC columns after stage 4, it is possible to assume that the infill walls and their openings
influenced negatively the seismic behaviour of the RC structure. The mortar rendering applied
to the RC frame was severely damaged after stage 3 (2475 YRP), specially at the corners of
the model. The single leaf infill walls with bed joint reinforcement connected to the RC frame
had a very good seismic performance, with no visible damage for the seismic standard
accelerations during stage 2 (475 YRP). After the last stage (4574 YRP) none of the infill
walls collapsed out-of-plane, even though the ones with openings at the ground floor
SERIES 227887 MASONRY ENCLOSURES Project
120
presented damage beyond repair. It was clear that the bed joint reinforcement prevented the
out-of-plane collapse due to its connection to the RC frame, otherwise the infill walls would
collapse as a rigid body.
Model 3 presented no considerable damage in the RC structure due to the accelerations
prescribed in the seismic standard during stage 2 (475 YRP), and after stage 3 (2475 YRP) the
damage was also light. After the retesting using the first three stages, with considerable loss
of stiffness, the model still presented visual light damage. After the reinforced rendering
removal, the RC columns did not present hinges at the extremities neither cracks at mid-
height, hence no undesirable collapse mechanism was developed. Given that model 3 was
designed following the EC2 and EC8 [29], [30], it is safe to say that the infill walls did not
influence undesirably the seismic behaviour of the RC structure. The infill walls presented
light damage after all the stages, even though the dynamic data presented a clear stiffness
loss. This was due to detachment of the infill wall from the reinforced rendering, allowing the
wall to move and not influence the behaviour of the RC structure, fact confirmed after the
rendering removal as the infill walls presented barely any cracks but were detached from the
RC structure. Although the reinforced rendering concealed the damage from the infill wall, it
also prevented the out-of-plane collapse because it was applied on both sides of the infill wall
and nailed to the RC frame and infill wall.
As for the comparison between models, some parameters are able to present differences
between the chosen reinforcement solutions of the infill walls, but a global trend that
highlights the better performance of one regarding others does not exist. It is clear that
model 1, designed using an older generation of standards [] and unreinforced solutions for the
infill walls had an undesirable seismic performance by collapsing, and the models designed
the Eurocodes [29], [30] did not, but along the test the amount of damage and global
behaviour were similar. This is due to the same geometry of the RC structure and infill walls
which are the parameters that most influence the structural behaviour, given that the strength
of the RC used was not significantly different. The detailing of the RC using [29] led to a
higher ductility in models 2 and 3 when compared to the detailing of model 1 using [35], and
the use of reinforcement in the infill walls attached to the RC frame prevented the
out-of-plane collapse, and possibly the collapse of the RC structure as well, but did not led to
less damage along the tests when compared to the unreinforced solution.
SERIES 227887 MASONRY ENCLOSURES Project
121
References
[1] Angel, R. Behavior of reinforced concrete frames with masonry infills. PhD Thesis, Dept. of Civil
Engineering, University of Illinois, Urbana-Champaign, USA, 1994.
[2] Bedell, R. J. and D. P. Abrams. Scale relationships of concrete columns. Research Report, Dept. of Civil
Engineering, Environment and Architecture, University of Colorado, Boulder, USA, 1983.
[3] Calvi, G. M., Bolognini, D. and A. Penna. Seismic performance of masonry-infilled r.c. frames: benefits
of slight reinforcements. Proc. of the 6th Portuguese Conference on Seismology and Earthquake
Engineering, Guimarães, Portugal, 2004.
[4] Calvi, G. M., Bolognini, D. and A. Penna. Design of masonry structures with bed joint reinforcement.
Proc. of the Seminar on masonry walls, Guimarães, Portugal, 2007.
[5] Carvalho, E. C. Seismic testing of structures. Proc. of the 11th European Conference on Earthquake
Engineering, Paris, France, 1998.
[6] Coelho, E. and F. Carvalho. Ensaios sísmicos – Instalações do LNEC. Journal Engenharia e Vida, No. 10,
2005, pp. 51-55.
[7] Cosenza, E. and G. Manfredi. Damage indices and damage measures. Progress in Structural Engineering
and Materials, Vol. 2, No. 2, 2000, pp. 50-59.
[8] Drysdale, R. G., A. A. Hamid, et al. Masonry Structures: Behavior and Design. Boulder, Colorado,
2008.
[9] Endevco. 7290A-10 characteristics. At https://www.endevco.com/7290a-10/ (accessed on 30/6/2013).
[10] Ewins, D. Modal testing: theory, practice and application. Research Studies Press, Hertfordshire, UK,
2000.
[11] Flanagan, R. D. and R. M. Bennett. Bidirectional behaviour of structural clay tile infilled frames.
Journal of Structural Engineering, Vol. 125, No. 3, 1999, pp. 236-244.
[12] Hamamatsu. C5949 characteristics. At http://pdf.datasheetcatalog.com/datasheet/hamamatsu/C5949.pdf
(accessed on 30/6/2013).
[13] He, J. and Z. F. Fu. Modal Analysis. Linacre House, Jordan Hill, Oxford, 2001.
[14] Jorquera, L. G. Estudio experimental sobre la resistencia de muros de albañileria sometidos a cargas
horizontales. Journal of IDIEM No. 3, 1964.
[15] Klingner, R. E. and V. V. Bertero. Infilled frames in earthquake-resistant construction. Research
Report, University of California, Berkeley, USA, 1976.
[16] Klingner, R. E. and V. V. Bertero. Earthquake resistance of infilled frames. Proc. of the American
Society of Civil Engineers, EUA, 1978.
[17] Komaraneni, S. Out-of-plane seismic behaviour of brick masonry infilled panels with prior in-plane
damage. PhD Thesis, Dept. of Civil Engineering, IIT Kanpur, India, 2009.
[18] Kramer, S. L. Geotechnical earthquake engineering. Prentice Hall, New Jersey, USA, 1996.
SERIES 227887 MASONRY ENCLOSURES Project
122
[19] Krypton Electronic Engineering n.v., K400/K600 - Hardware & Software Guide (Krypton Help Pages),
Leuven, 2003.
[20] Leite, J. Avaliação da Segurança Baseada em Deslocamentos: Aplicação a Edifícios de Betão Armado
sem e com Paredes de Preenchimento. MSc Dissertation, Dept. of Civil Engineering, University of
Minho, Guimarães, Portugal, 2009.
[21] Lemaitre, J. and R. Desmorat. Engineering damage mechanics: Ductile, creep, fatigue and brittle
failures. Springer, 2005.
[22] LNEC. Main characteristics of triaxial shaking table. At
http://www.lnec.pt/organizacao/de/nesde/ptriaxialcaracteristicas (accessed on 30/6/2013).
[23] Mainstone, R. J. On the stiffness and strength of infilled frames. Proc. of the Institution of Civil
Engineers, No. 49, London, UK, 1971.
[24] Mallick, D. V. and R. P. Garg. Effect of openings on the lateral stiffness of infilled frames. Proc. of the
Institution of Civil Engineers, No. 49, London, UK, 1971.
[25] Mehrabi, A. B. and P. B. Shing. Performance of masonry-infilled r/c frames under in-plane lateral
loads. Research Report, Dept. of Civil Engineering, Environment and Architecture, University of Colorado,
Boulder, USA, 1994.
[26] Mendes, L. LNEC-SPA: Signal Processing and Analysis Tools for Civil Engineers. Earthquake
Engineering and Structural Dynamics Division, Dept. of Structures, National Laboratory for Civil
Engineering (LNEC), Lisbon, Portugal, 2008.
[27] Miranda, E. and S. Taghavi. Approximate floor acceleration demands in multistory buildings. I:
Formulation. Journal of Structural Engineering, Vol. 131, No. 2, 2005, pp. 203-211.
[28] Miranda, E. and S. D. Akkar. Generalized interstory drift spectrum. Journal of Structural Engineering,
Vol. 132, No. 6, 2006, pp. 840-852.
[29] NP EN 1992-1-1. Eurocódigo 2 - Projecto de estruturas de betão - Parte 1-1: Regras gerais e regras
para edifícios. Portuguese Institute for Quality (IPQ), Lisbon, Portugal, 2010.
[30] NP EN 1998-1. Eurocódigo 8 - Projecto de estruturas para resistência aos sismos - Parte 1: Regras
gerais, acções sísmicas e regras para edifícios. Portuguese Institute for Quality (IPQ), Lisbon, Portugal,
2010.
[31] PCB Piezotronics. 393A03 characteristics. At http://www.pcb.com/products.aspx?s=393A03 (accessed on
30/6/2013).
[32] PCB Piezotronics. 626B13 characteristics. At http://www.pcb.com/products.aspx?s=626B13 (accessed on
30/6/2013).
[33] Penna, A., Lagomarsino, S. and A. Galasco. A nonlinear macroelement model for the seismic analysis
of masonry buildings. Earthquake Engineering and Structural Dynamics, Vol. 43, No. 2, 2013, pp. 159-
179.
[34] Pereira, M. F. P. Avaliação do desempenho das envolventes dos edifícios face à acção dos sismos. PhD
Thesis, Dept. of Civil Engineering, University of Minho, Guimarães, Portugal, 2013.
[35] REBAP. Regulamento de Estruturas de Betão Armado e Pré-Esforçado. Diário da República, Portugal,
1983.
[36] RSA. Regulamento de Segurança e Acções para Estruturas de Edifícios e Pontes. Diário da República,
Portugal, 1983.
[37] SAP2000. Structural Analysis Program, Version 15.0.0. Computers and Structures, Inc. Berkeley, EUA,
2013.
SERIES 227887 MASONRY ENCLOSURES Project
123
[38] Seismosoft. SeismoStruct: A computer program for static and dynamic nonlinear analysis of framed
structures, Version 6.9.5. Available at http://www.seismosoft.com, Pavia, Italy, 2013.
[39] Stafford Smith, B. Model test results of vertical and horizontal loading of infilled frames. ACI Journal,
No. 65, 1968.
[40] Wilcoxon. 799M characteristics. At
http://www.wilcoxon.com/vi_index.cfm?PD_ID=14&P_ID=14&search=y (accessed on 30/6/2013).
[41] Zarnic, R. and M. Tomazevic. Study of the behaviour of masonry infilled reinforced concrete frames
subjected to seismic loading - Part II. Research Report, University of Ljubljana, Slovenia, 1985.