STUDY AND DEVELOPMENT

OF A HIGH PERFORMANCE
SHAKE TABLE SYSTEM - TREES LAB 2

A Dissertation Submitted in Partial Fulfilment of the Requirements

for the Master Degree in

Earthquake Engineering

By

Alessandro Scodeggio

Supervisor(s): Prof. Alberto Pavese, Dott. Ing. Filippo Dacarro

May, 2011

Istituto Universitario di Studi Superiori di Pavia

Università degli Studi di Pavia
The dissertation entitled “Study and development of a high performance shake table system -
TREES Lab 2”, by Alessandro Scodeggio, has been approved in partial fulfilment of the
requirements for the Master Degree in Earthquake Engineering.

Name of Reviewer 1 …… … ………

Name of Reviewer 2………… … ……

Abstract

ABSTRACT

A wide area of the European continent can be considered as a high seismicity zone. The
number of human and economic losses under seismic events is significant comparing with the
losses in countries like United States and Japan which are more often affected by earthquakes.
Even if the number of human losses is decreasing in the most industrialized countries, the
sustaining costs after a significant earthquake are steadily increasing. For this reason it’s
really important improve the research in this engineering field. The recent event occurred in
L’Aquila (Italy) showed the high risk for the existing buildings, where the coexistence
between old and new structure leads to a worse situation in terms of losses.

The structural behaviour to seismic loading is fundamental not only in civil engineering but
also in the field of mechanical engineering. The applications with cyclic dynamic inputs are
difficult to model and so the monitoring of the existing building is nowadays not
economically profitable. For these reasons the laboratory tests have a really important role in
understanding the seismic behaviour of structures and improving both design and assessment
procedures. A seismic simulation allows the evaluation of the seismic response of the
structure and of the structural components in order to verify the performance of it. Nowadays
seismic tests are increasingly used in the field of engineering and several high performance
shake table systems are operating around the world. The most updated facilities are equipped
with hydraulic pumps, accumulators, servovalves and controlling system to allow the right
reproduction of acceleration records and synthetic time-histories with high demanding
frequency contents. The problem for all the laboratories is clearly the precision with which the
real conditions are recreated. The new laboratory TREES Lab 2 will be design in order to
improve the actual capacity of the existing facility at EUCENTRE by considering all the most
important requirements to construct an efficient and powerful laboratory.

Keywords: losses; risk; earthquakes; simulation; response; tests; performances; laboratory.

i
Acknowledgements

ad Agata e Francesco

.....un grazie speciale a Danae, Reyes, Soccer, Andres, Mirko, Nicola, Costas, Ricardo, Alexandra,
Alexis, Dago, Lenore, Susie, Robin e Marco per essermi stati amici e compagni di questa unica e
bellissima esperienza...ringrazio anche Filippo per avermi aiutato a concludere questo percorso con
la sua disponibilità e competenza.....

ii
Index

TABLE OF CONTENTS

Page
ABSTRACT...........................................................................................................................................i
TABLE OF CONTENTS.....................................................................................................................iii
LIST OF FIGURES..............................................................................................................................vi
LIST OF TABLES................................................................................................................................x
1 NEEDS FOR ADVANCED EXPERIMENTAL FACILITIES........................................................1
2 BRIEF REVIEW OF THE EXISTING FACILITIES......................................................................6
2.1 USA - San Diego (LHPOST)....................................................................................................6
2.2 USA - Buffalo (SEESL)...........................................................................................................9
2.3 USA - Reno............................................................................................................................10
2.4 JAPAN - Miki-city (E-DEFENSE).........................................................................................13
2.5 TAIWAN - Taipei (NCREE)..................................................................................................14
2.6 UK - Bristol (EQUALS).........................................................................................................16
2.7 PORTUGAL - Lisbon (LNEC-3G)........................................................................................17
2.8 FRANCE - Saclay (AZALEE)...............................................................................................18
2.9 ITALY - Pavia (TREES Lab).................................................................................................19
2.10 Comparisons...........................................................................................................................21
3 INTRODUCTION OF THE NEW TREES LAB 2 AT EUCENTRE.............................................22
3.1 Location..................................................................................................................................23
3.2 Objectives and applications....................................................................................................23
3.3 Requirements of the facility....................................................................................................24
3.4 Local needs and constraints....................................................................................................25
3.5 General description of the new dynamic testing facility specifications..................................25
4 COMPONENT OF THE SHAKE TABLE FACILITY..................................................................28
4.1 Actuators................................................................................................................................28

iii
Index

4.1.1 Force.............................................................................................................................29
4.1.2 Specifications of the actuator........................................................................................30
4.1.3 Swivel rod end..............................................................................................................32
4.1.4 Spiral washer................................................................................................................32
4.1.5 Force transducers..........................................................................................................32
4.1.6 LVDT...........................................................................................................................32
4.1.7 Swivel bases.................................................................................................................32
4.1.8 Typical actuator configurations....................................................................................35
4.1.9 Flow..............................................................................................................................36
4.2 Bearings..................................................................................................................................36
4.3 Hold-Downs...........................................................................................................................37
4.4 Servovalves............................................................................................................................38
4.5 Oil column resonance.............................................................................................................44
4.6 Lateral resonance effects (bow string frequency)...................................................................44
4.7 Accumulators..........................................................................................................................45
4.7.1 Bladder accumulators...................................................................................................45
4.7.2 Piston accumulators......................................................................................................47
4.7.3 Conventional accumulator system................................................................................48
4.7.4 Blow down accumulator system...................................................................................49
4.7.5 Pumps versus accumulator blow down as a function of time........................................49
4.7.6 Horizontal half-sine shock on 120 tons specimen.........................................................49
4.8 Hydraulic pumps....................................................................................................................50
4.8.1 Diesel pumps................................................................................................................50
4.8.2 Electric pumps..............................................................................................................51
4.9 Cooling system.......................................................................................................................54
5 LAYOUT OF THE SHAKE TABLE SYSTEM............................................................................55
5.1 First draft................................................................................................................................56
5.2 Second draft............................................................................................................................57
6 COMPUTATION OF THE SYSTEM............................................................................................61
6.1 Description of the numerical evaluations................................................................................61
6.1.1 Displacement actuator demand.....................................................................................63
6.1.2 Force.............................................................................................................................77
6.1.3 Hydraulic system (flow)...............................................................................................78
6.1.4 Accumulators demand..................................................................................................80
6.2 Specifications for the shaking table design.............................................................................87

iv
Index

6.2.1 Dynamic response characteristics.................................................................................88

6.2.2 Overturning moment...................................................................................................100
6.3 Specifications for the system restraining the platform motion..............................................101
6.4 Specifications for the reaction mass design..........................................................................101
6.5 Specifications for the isolation and damping system design.................................................102
7 CONCLUSIONS..........................................................................................................................103
8 REFERENCES.............................................................................................................................104
APPENDIX A.......................................................................................................................................1

v
Index

LIST OF FIGURES

Page
Figure 1 - Testing set up of a full scale building at E-DEFENSE..............................................1
Figure 2 - View of both shake tables at SEESL..........................................................................2
Figure 3 - View of the shake table B at SEESL..........................................................................2
Figure 4 - TREES Lab (EUCENTRE) - Actuator of the SDOF shake table system and
reaction wall........................................................................................................................3
Figure 5 - Three-storey infilled RC frame tested at San Diego (LHPOST)...............................4
Figure 6 - Seven-storey building tested at San Diego (LHPOST)..............................................4
Figure 7 - NEES Large High Performance (LHP) Outdoor Shake Table...................................7
Figure 8 - NEES Large High Performance (LHP) Outdoor Shake Table - Sinusoidal
performance estimate..........................................................................................................8
Figure 9 - Layout of the shake table system at USCD (LHPOST) - Upper view.......................8
Figure 10 - Layout of the shake table system at UCSD (LHPOST) - Bottom view...................8
Figure 11 - NEES at the University of Nevada - Curved bridge experiment proposal using 4
tables.................................................................................................................................12
Figure 12 - E-DEFENSE - Construction phase........................................................................13
Figure 13 - NCREE - View of the laboratory during dynamic testing.....................................15
Figure 14 - Arrangement of the three-axis table at LNEC [ECOEST. 1997]...........................17
Figure 15 - AZALEE - View of the laboratory.........................................................................18
Figure 16 - TREES Lab - View of the laboratory.....................................................................20
Figure 17 - Position of Eucentre (Pavia, Italy).........................................................................22
Figure 18 - Prospectus of the laboratories with inner view......................................................24
Figure 19 - Single acting actuator.............................................................................................28
Figure 20 - Double acting actuator - Single rod.......................................................................28

vi
Index

Figure 21 - Double acting actuator - Double rod......................................................................28

Figure 22 - Cutaway view of MTS Series 244 hydraulic actuator...........................................30
Figure 23 - Channel system......................................................................................................31
Figure 24 - Actuator series 244.................................................................................................32
Figure 25 - Characteristics of the selected actuator..................................................................34
Figure 26 - Characteristics of the actuator swivel....................................................................34
Figure 27 - LVDT specifications..............................................................................................35
Figure 28 - Parts of the selected Hold-Down............................................................................37
Figure 29 - Characteristics of the selected Hold-Downs..........................................................37
Figure 30 - Characteristics of the Hold-Down swivel..............................................................37
Figure 31 - Cutaway view of a 256 servovalve........................................................................38
Figure 32 - Series 256 servovalve.............................................................................................39
Figure 33 - Typical closed-coop servo-hydraulic testing System.............................................40
Figure 34 - Servovalve flow-frequency performance curve 256.40A......................................41
Figure 35 - Servovalve flow-frequency performance curve 256.40B......................................41
Figure 36 - Servovalve dimensions...........................................................................................42
Figure 37 - Bladder accumulator schematic.............................................................................46
Figure 38 - Bladder accumulator..............................................................................................46
Figure 39 - Open view of a floating piston accumulator..........................................................47
Figure 40 - Blow down accumulator bank system - MOC.......................................................49
Figure 41 - Cross sectional view of a vane pump.....................................................................52
Figure 42 - Cross sectional view of a radial-piston pump........................................................52
Figure 43 - Cross sectional view of an axial piston swash plate pump....................................53
Figure 44 - TREES Lab 2 - First draft layout scheme..............................................................56
Figure 45 - TREES Lab 2 - First draft layout scheme..............................................................56
Figure 46 - Shake table layout - Position 0 of the actuators.....................................................58
Figure 47 - Shake table layout - Displacement along one of the two horizontal directions.....58
Figure 48 - Shake table layout - Combined displacement on the horizontal plane..................59
Figure 49 - Shake table layout - Rotation about the vertical axis.............................................59
Figure 50 - Shake table layout - Position 0 in the configuration at 1 DOF..............................60
Figure 51 - Shake table layout - Displacement along the horizontal direction for the SDOF
system configuration.........................................................................................................60
Figure 52 - Zero initial condition of the adopted shake table system configuration................62

vii
Index

Figure 53 - Global reference system for the calculation of the actuator displacement.............64
Figure 54 - Displacement of the shake table.............................................................................65
Figure 55 - Rotation of the shake table.....................................................................................66
Figure 56 - Combination of displacement and rotation for the shake table system..................67
Figure 57 - Displacement for actuator Y1................................................................................69
Figure 58 - Displacement for actuator Y2................................................................................69
Figure 59 - Displacement for actuator Y3................................................................................71
Figure 60 - Displacement for actuator Y4................................................................................71
Figure 61 - Displacement for actuator X1................................................................................73
Figure 62 - Displacement for actuator X2................................................................................73
Figure 63 - Displacement for actuator X3................................................................................75
Figure 64 - Displacement for actuator X4................................................................................75
Figure 65 - Rotation of the actuators in Y direction with no rotational input for the system
(red line)............................................................................................................................76
Figure 66 - Rotation of the actuators in X direction with no rotational input for the system
(red line)............................................................................................................................76
Figure 67 - Force for the component X and Y of the motion...................................................77
Figure 68 - Flow rate of the actuators in Y direction................................................................78
Figure 69 - Flow rate of the actuators in X direction................................................................79
Figure 70 - Total flow rate comparison....................................................................................79
Figure 71 - Shifted flow rate comparison.................................................................................80
Figure 72 - Total flow volume comparison..............................................................................81
Figure 73 - Acceleration records for the X direction................................................................83
Figure 74 - Acceleration records for the Y direction................................................................83
Figure 75 - Spectra of the acceleration records in the X direction...........................................84
Figure 76 - Spectra of the acceleration records in the Y direction...........................................84
Figure 77 - Total flow rate demand for the set of acceleration records....................................85
Figure 78 - Total flow volume demand for the set of acceleration records..............................86
Figure 79 - Flow volume demand for the accumulation system and accumulator capacity.....86
Figure 80 - Vertical alignments (longitudinal and transversal) x-y plane (measures in cm) -
First draft...........................................................................................................................88
Figure 81 - Horizontal alignments y-z plane (measures in cm) - First draft.............................90
Figure 82 - Horizontal alignments x-z plane (measures in cm) - First draft.............................90

viii
Index

Figure 83 - View of the shake table model and restraints (first draft)......................................90
Figure 84 - First mode shape 104 Hz (first draft).....................................................................91
Figure 85 - Second mode shape 109 Hz (first draft).................................................................92
Figure 86 - Third mode shape 166 Hz (first draft)....................................................................92
Figure 87 - Fourth mode shape 72 Hz (first draft)....................................................................93
Figure 88 - Vertical alignments (longitudinal and transversal) x-y plane (measures in cm) -
Second draft......................................................................................................................94
Figure 89 - Horizontal alignments y-z plane (measures in cm) - Second draft........................95
Figure 90 - Horizontal alignments x-z plane (measures in cm) - Second draft........................95
Figure 91 - View of the shake table model and restraints (second draft).................................95
Figure 92 - View of the inner layer scheme (second draft)......................................................96
Figure 93 - First mode shape 136 Hz (second draft)................................................................97
Figure 94 - Second mode shape 145 Hz (second draft)............................................................97
Figure 95 - Third mode shape 158 Hz (second draft)...............................................................98
Figure 96 - Fourth mode shape 159 Hz (second draft).............................................................98
Figure 97 - Position of the centre of gravity and position of the layers (measures in cm).....100
Figure 98 - Components of a typical shake table (plan view) [Clark, 1992]..........................102
Figure 99 - Components of a typical shake table (section) [Clark, 1992]..............................102
Figure 100 - Plan view of the two laboratories at EUCENTRE.................................................1
Figure 101 - Section A_A of TREES Lab 2 structure................................................................2
Figure 102 - Section B_B of TREES Lab 2 structure.................................................................2
Figure 103 - Plan view of the laboratories and the buildings of University of Pavia.................3

ix
Index

LIST OF TABLES

Page
Table 1 - General specifications of NEES Large High Performance (LHP) Outdoor Shake
Table...................................................................................................................................7
Table 2 - SEESL Shake Tables A & B specifications................................................................9
Table 3 - SEESL Shake Tables O specifications......................................................................10
Table 4 - NEES at the University of Nevada, Reno specifications..........................................11
Table 5 - E-Defense specifications...........................................................................................14
Table 6 - NCREE specifications...............................................................................................15
Table 7 - TREES Lab specifications.........................................................................................20
Table 8 - Comparison of the performances for the considered shake table systems................21
Table 9 - General characteristics of the new shaking table at TREES Lab 2...........................27
Table 10 - LVDT specifications...............................................................................................33
Table 11 - Static servovalve performance................................................................................40
Table 12 - Dynamic servovalve performance...........................................................................41
Table 13 - Servovalve dimensions table...................................................................................43
Table 14 - Comparison between first and second draft............................................................57
Table 15 - Set of earthquake records for the flow requirements calculations..........................82
Table 16 - Thickness of the shake table plates (first draft).......................................................89
Table 17 - Result of the modal analysis (first draft).................................................................91
Table 18 - Participating masses (first draft)..............................................................................93
Table 19 - Result of the modal analysis (second draft)............................................................96
Table 20 - Participating masses (second configuration)...........................................................99
Chapter 1. Needs for Advanced Experimental Facilities

1 NEEDS FOR ADVANCED EXPERIMENTAL FACILITIES

In the last decades the research results in the field of earthquake engineering have contributed
to develop a better understanding of the problems related to the structural behaviour leading
to the preparation of the modern codes and to the identification of new design and assessment
procedures. Predictive models have frequently calibrated on experimental results and the
validation of the behaviour of large structure requires advanced testing facilities with high
performances in terms of dimension, payload, system DOFs, peak acceleration, peak velocity,
peak displacement and frequency range in order to avoid scaling effects.

Figure 1 - Testing set up of a full scale building at E-DEFENSE

Chapter 1. Needs for Advanced Experimental Facilities

Dimensions and payload allow testing of full scale building while all the others aspects deal
with the possibility of reproducing acceleration records with no limits in terms of peak
acceleration, peak velocity, peak displacement and frequency content. A testing system with
those characteristics is an indispensable tool to calibrate and validate new conceptual
approaches in modelling and simulations developed for performance based analysis and
design of new structures or retrofitting interventions of safe structures.

Shake table testing is one of the most widespread experimental method in earthquake
engineering since it allows to reproduce directly an acceleration input. Shake table tests are
real-time dynamic tests and they can deal with phenomena that cannot be accounted for by
other testing methods. They are more representative of real behaviour in case of rate-
dependent constitutive laws. Regarding the DOFs of the system, it’s possible to apply
horizontal and vertical seismic excitations and rotational inputs. It’s also feasible to account
for dynamic coupling between horizontal excitation and vertical response (e.g. rocking and
overturning of solids, dynamic variation of axial force due to cracking of reinforced concrete
walls subjected to horizontal excitation).

Figure 2 - View of both shake tables at SEESL Figure 3 - View of the shake table B at SEESL

New facilities should have the capability of applying high intensity excitations to model that
will be representative of the prototype structures. Because of all the advantages given by
using this testing system, potential users of earthquake experimental facilities are research
laboratories, industries and construction companies.
Chapter 1. Needs for Advanced Experimental Facilities

Figure 4 - TREES Lab (EUCENTRE) - Actuator of the SDOF shake table system and reaction wall

According to their main objectives, seismic tests can be classified in three categories. The first
type deals with qualification tests that allow the qualification of the seismic response of
equipments. The scale of the test, in this case, have to be 1:1 in order to avoid scaling effect
since the prototype has to be qualified. In the second category, then, it’s possible to find the
demonstration tests which deal with strategic reasons such as to convince policy makers.
Generally, the tested models are at the biggest possible scale. Finally, in the third category
there are the research and development oriented tests which are necessary for a better
understanding of the structural behaviour, at the local and global level, in order to propose or
validate codes recommendation and guidelines and to validate or calibrate numerical models.
Chapter 1. Needs for Advanced Experimental Facilities

Figure 5 - Three-storey infilled RC frame tested at San Diego (LHPOST)

For all the categories previously mentioned, a high performance shake table system is
necessary. The system involves mainly, apart from the shake table itself, the hydraulic system
able to reproduce the input motions, the control system to avoid considerable discrepancies
between target motion and achieved motion and the measurement system necessary to enable
a satisfactory exploration of the experimental results.

Figure 6 - Seven-storey building tested at San Diego (LHPOST)

It’s also important to mention that a strong interaction between an experimental facility and a
high computational capability facility, on site or remote, is necessary. Nowadays, high
performance and accuracy numerical simulation is necessary not only for advanced
experimental methods, like real-time sub-structuring involving complex numerical
substructures, but also for conventional tests.
Chapter 1. Needs for Advanced Experimental Facilities

In order to have a successful testing of models with complex behaviour, it’s really important
being able to obtain accurate results of predictive analyses before and after the testing.
Predictive analyses are necessary to define the testing configuration. On the other hand,
interpretation analyses may be useful for the detection of problems or unexpected response
that occurs during the test.

Finally, considering the distribution of high performance seismic testing facilities all around
the world, it has to be noticed that several of those facilities are either operating or under
construction outside Europe, mainly in USA, Japan and China. In order to have a better
understanding of what the specifications of a new high performance shake table system
should be, a description of some of the most important facilities around the world can be
found in the next chapter.
Chapter 2. Brief Review of the Existing Facilities

2 BRIEF REVIEW OF THE EXISTING FACILITIES

This work will provide in the following pages a relatively detailed description of the design
process for the new shake table system at TREES Lab 2. Before going through the design part,
a short qualitative description of the most important shake table facilities available all over the
world is given. Considering the European facilities, TREES Lab 2 will definitely increase the
potentiality of the continental net of earthquake simulators with its performances.

2.1 USA - San Diego (LHPOST)

The UCSD Large High Performance (LHP) Outdoor Shaking Table is located at the Field
Station at Camp Elliott, a site at 15 km away from the main UCSD campus. The NEES at
UCSD LHPOST is part of the Englekirk Structural Engineering Centre (ESEC). In addition to
the NEES at UCSD shake table, the Englekirk Centre houses a Soil-Foundation-Structure
Interaction (SFSI) facility, funded by the California Department of Transportation (Caltrans)
and a Blast Simulator, funded by the Technical Support Working Group (TSWG). The NEES
at UCSD Large High Performance Outdoor Shake Table is capable of creating realistic
simulations of the most devastating earthquakes and has no height restrictions.

The Network for Earthquake Engineering Simulation (NEES) LHP Outdoor Shaking Table is
a 7.6 m wide by 12.2 m long, SDOF system with the capability of being upgraded to 6 DOFs.
With respect to the maximum capabilities of the system, the specimen weight reaches 400
tons even if the vertical payload capacity is 2000 tons. The specifications for the first phase of
the facility are a peak acceleration of 4.7 g for a bare table, a peak horizontal velocity of 1.8
m/s and a stroke of ± 750 mm. The testing frequency range is 0-20 Hz. Although this table is
not the largest of its kind in terms of size in the world, the velocity, frequency range, and
stroke capabilities make it the world's first outdoor shaking table. The intention of NEES was
to add a significant new dimension and capabilities to existing United States testing facilities
with no overhead space and lifting constraints. The outdoor table may also contribute in
overcoming current problems with damage incurred during transportation of specimens.
Chapter 2. Brief Review of the Existing Facilities

Figure 7 - NEES Large High Performance (LHP) Outdoor Shake Table

Specifications for LHPOST

Dimensions 7.6x12.2 m
Mass -
Payload 20 MN
DOFs 1
Maximum force capacity 6.8 MN
Stroke +/- 750 mm
Peak velocity +/- 1.8 m/sec
Peak acceleration: bare table, 400 tons payload 4.2 g, 1.2 g
Maximum overturning moment: bare table, 400
35 MNm, 55 MNm
ton payload
Yaw rotation -
Operation frequency 0-20 Hz
Table 1 - General specifications of NEES Large High Performance (LHP) Outdoor Shake Table
Chapter 2. Brief Review of the Existing Facilities

Figure 8 - NEES Large High Performance (LHP) Outdoor Shake Table - Sinusoidal performance estimate

Finally, the equipment of the shaking table includes:

- 2 HPU’s unit (SRMD) – 120 gpm at 5000 psi and 180 gpm at 3000 psi;
- 50-50 gallon accumulators capable of supplying approximately 680 gallons of oil (or 7.5
m swept displacement) at 3000 psi;
- 2 MTS Model 243 longitudinal actuators;
- 6 MTS Model 270 style vertical pressure balanced bearings;
- MTS model 469D digital control system, electronic hardware;
- 2 nitrogen Hold-Downs struts;
- hydraulic system with 9500 liter accumulator bank and 19000 liter surge tank;
- platen design including lateral restraint bearings;
- 2 servovalves per actuator and 10000 liter/min flow rating (4th stage);
- 6 MTS Model 270 style vertical pressure balance bearings.

Figure 9 - Layout of the shake table system at Figure 10 - Layout of the shake table system at
USCD (LHPOST) - Upper view UCSD (LHPOST) - Bottom view
Chapter 2. Brief Review of the Existing Facilities

2.2 USA - Buffalo (SEESL)

The Structural Engineering and Earthquake Simulation Laboratory (SEESL) was created in
1983 and was founded on its’ predecessor, the Structural Engineering Laboratory (SEL) at the
University at Buffalo. This new laboratory is equipped with an advanced 5-degree-of-
freedom 50 tons capacity shake table.

In 2004 SEESL completed an expansion which added two new relocatable 6 DOFs 50 tons
shake tables, advanced large scale dynamic and static actuators, a geotechnical laminar box,
and networked controllers interfaced to a high speed local area network.

The SEESL laboratory equipment includes:

- a set of two high-performance, 6 DOFs shake tables, which can be rapidly repositioned
from directly adjacent to one another to positions up to 30 m apart (centre-to-centre);
- together, the tables can host specimens of up to 100 tons and as long as 35 m, and subject
them to fully in-phase or totally uncorrelated dynamic excitations;
- next to the shake tables, a large reaction wall equipped with external dynamic actuators
allows application of computer-simulated forces on shake table-mounted substructures,
thus simulating the reaction of the entire structure (Real-Time Dynamic Hybrid Testing,
or RTDHT);
- equipment required to operate the shake tables, including a high-performance hydraulic
supply and distribution system (up to 1600 gpm), and numerous digital control systems;
- networked tele-experimentation capabilities using modular and expandable tele-
observation and tele-operation equipment, tied to the testing systems using discrete and
global sensors, including high-resolution digital video and imaging capabilities;
- the 3.66 by 3.66 m shake table has 6 controlled DOFs. This shake table has been in use at
the University at Buffalo for nearly 20 years. In 2004, it has been refurbished with a new
controller and re-built actuators.

Specifications for SEESL - Shake tables A&B

Dimensions (with extension platform in place) 3.6x3.6 m (7x7 m)
Mass -
Payload (nominal) 50 tons (20 tons)
Payload with table extension platform in place 40 tons
DOFs 6
Maximum force capacity -
Stroke (X, Y and Z direction) +/- 150 mm, +/- 150, +/- 75 mm
Peak velocity (X, Y, Z direction) 1.25 m/sec, 1.25 m/sec, 0.5 m/sec
Peak acceleration (X, Y and Z direction) 1.15 g, 1.15 g, 1.15 g (with 20 tons specimen)
Maximum overturning moment 46 ton m
Yaw rotation -
Operation frequency 0-50 Hz nominal, 100 Hz maximum
Chapter 2. Brief Review of the Existing Facilities

Table 2 - SEESL Shake Tables A & B specifications

Chapter 2. Brief Review of the Existing Facilities

Specifications for SEESL – Shake table O

Dimensions 3.66x3.66 m
Mass -
Payload (nominal) 50 tons (20 tons)
DOFs 5
Maximum force capacity -
Stroke (X and Z direction) +/- 150 mm, +/- 75 mm
Peak velocity (X and Z direction) 0.762 m/sec, 0.5 m/sec
Peak acceleration (X and Z direction) +/- 1.15 g, +/-2.30 g (with 20 tons specimen)
Maximum overturning moment 46 ton m
Yaw rotation -
Operation frequency 0-50 Hz nominal
Table 3 - SEESL Shake Tables O specifications

2.3 USA - Reno

The NEES Equipment Site at the University of Nevada, Reno is a multiple-shake-table facility
with three identical biaxial and one 6 DOFs shake tables. The facility is also capable of testing
conventional structural and non-structural systems by using the biaxial tables in large-table-
mode, and operating them as a single unit.

The Laboratory has three 50 tons biaxial shake tables in operation. Manufactured by MTS,
each table measures 4.3 by 4.5 m, has a stroke of +/- 300 mm, and can reach a peak velocity
of 1000 mm/sec and an acceleration of 1 g under the full 50 tons payload. All there tables are
relocatable. To achieve these maxima, two banks of blow-down accumulators are used. With
the actuator under the table, the tables are designed to handle simultaneously a maximum
pitch moment of 1356 kNm along with a maximum roll and yaw moment of 542 kNm. If a
single table were to be operated by the continuous oil supply, the maximum velocity reduces
to 635 mm/sec. All three tables can be constrained to act together as a single large table, or
they can be operated individually with independent motions.
Chapter 2. Brief Review of the Existing Facilities

Specifications for University of Nevada Shake Table

Dimensions 4.3x4.5 m
Mass -
Payload 45 tons
DOFs 6
Maximum force capacity 734 kN
Stroke +/- 350 mm static, +/- 300 mm dynamic
Peak velocity 1.27 m/sec
Peak acceleration 1g
Maximum pitch moment 1350 kNm
Maximum yaw moment 542 kNm
Maximum roll moment 734 kNm
Yaw rotation -
Operation frequency 0-50 Hz nominal
Table 4 - NEES at the University of Nevada, Reno specifications

NEES at University of Nevada has recently added a fourth shake table referred as the 6 DOFs
table to complement its existing three biaxial shake tables. The fourth table is a true 6 DOFs
system with three translational and three rotational components.

Three hydraulic power units (MTS 506.82, MTS 506.62, and MTS 506.41) provide the
pressurized hydraulic fluid to the hydraulic distribution system in direct response to control
signals received from the Remote Pump Control Panel. The hydraulic service manifolds use
check valves, pressure relief valves, accumulators, and solenoid controlled valves to adjust the
hydraulic pressure to the MTS multi-stage servovalves and blowdown system. The blowdown
accumulators are used to maintain system performance during short-term, high-demand test
conditions. There is one blowdown system for each actuator. Flow averaging accumulators
mounted on each leg of the main pressure piping and the main return piping reduce the
pressure fluctuations in the hydraulic system. A sump pump returns hydraulic fluid that has
accumulated in the sump reservoir to the hydraulic fluid reservoir through the main return
pipe.

The actuators (series 244.50S actuators) provided for the seismic test system are specially
designed for high performance and they have a dynamic stroke of 300 mm. An internally
mounted actuator is located under the table and is attached to the table and internal buttress. A
linear variable differential transformer (LVDT) is mounted inside the actuator to provide an
electrical feedback signal that indicates the actuator position. The actuators are equipped with
multistage servovalves which control the direction and amount of fluid flow to the actuators.
The actuators have two Series 256.25 servovalves which provide a total rated flow of 1890
lpm.
Chapter 2. Brief Review of the Existing Facilities

The table is comprised of a 3 by 3 m, 7 tons platen and can carry a nominal payload of 20 tons
at 1 g acceleration in all axes with a maximum payload of 60 tons at reduced accelerations.
The X, Y and Z displacement strokes are nominally 150, 610 and 200 mm. The system uses 7
MTS actuators and an MTS 469D seismic controller, which is integrated into the laboratory
computer network through ScramNET reflective memory. The 6 DOFs table is re-locatable
and can be operated synchronously or asynchronously with the current three shake table bi-
axial system. The table was added to the NEES at University of Nevada equipment inventory
in 2008.

Figure 11 - NEES at the University of Nevada - Curved bridge experiment proposal using 4 tables
Chapter 2. Brief Review of the Existing Facilities

2.4 JAPAN - Miki-city (E-DEFENSE)

The 3-D Full Scale Earthquake Testing Facility (E-Defense) is located in the Miki Earthquake
Disaster Memorial Park, constructed, north of Kobe-city, by the Hyogo Prefecture
Government. This is the world's largest facility for 3-dimensional vibration testing.

The table size is 20 by 15 m. The payload reaches 1200 tons. The maximum acceleration is
1.7 g with zero payload, the table velocity has a peak of 2 m/s and the maximum displacement
is ± 1000 mm. High-capacity servo-hydraulic equipment is required in order to drive the large
mass of the table and test specimen at the required rate. The very large E-Defense table
required the development of new bearing and pressure seal systems as well as servovalves
able to provide flows of 15000 l/min. Three servovalves are provided for each horizontal
actuator and one valve for each vertical actuator. This means a total flow rate of 45000 l/min
is needed for one horizontal actuator in order to realize a table velocity of 2 m/s.

The actuators of the shake table are designed and made on the basis of moving the table in the
frequency range of 1 to 15 Hz. The rated thrusts total 2300 tons each in the horizontal X and
Y directions and 8300 tons in the vertical Z direction. The entire reaction force is transferred
to the reinforced concrete foundations. In order to ensure the performance of the foundations,
the mass of the foundations was required to be around 200000 tons, the thickness of the
bottom block that supports the vertical actuators at least 7 m, and the thickness of the wall that
support the horizontal actuators in the X and Y directions at least 13 m.

Figure 12 - E-DEFENSE - Construction phase

Chapter 2. Brief Review of the Existing Facilities

Specifications for E-Defense

Dimensions 20x15 m
Mass 750 tons
Payload 1200 tons
DOFs 3
Maximum force capacity -
Stroke (X, Y, Z axis) +/- 1000 mm, +/-1000 mm +/- 500 mm
Peak velocity (X, Y, Z axis) 2 m/sec, 2 m/sec, 0.7 m/sec
Peak acceleration (X, Y, Z axis) 9 m/sec2, 9 m/sec2, 15 m/sec2
Maximum overturning moment 1500 MN m
Maximum yaw moment 40 MNm
Yaw rotation -
Operation frequency 0-15 Hz nominal
Table 5 - E-Defense specifications

2.5 TAIWAN - Taipei (NCREE)

Through a joint effort of the National Science Council (NSC) and the National Taiwan
University (NTU), the National Centre for Research on Earthquake Engineering (NCREE)
was officially established in 1990. The seismic simulator in the NCREE possesses 6 DOFs
and the size of the shaking table of the seismic simulator is 5 m by 5 m with a mass of 27
tons. Structural model with maximum payload 50 tons can be accommodated on the table.
The shaking table is a box shaped structure so that high bending and torsional stiffness can be
achieved with limited mass of the table.

The shaking table is driven by 12 hydraulic actuators 4 actuators for each axis. The hydraulic
power is provided by 2 electrical pumps and 3 diesel pumps which offer a total flow rate of
1235 gpm with a working pressure of 210 kg/cm 2. The weight of the shaking table and the
structural model is balanced by 4 static supports. The reaction forces of the actuators are
provided by the reaction mass which is 16 m (length) by 16 m (width) and 7.6 m height with a
mass of 4000 tons. In order to further improve the quality of the testing environment, the
reaction mass is isolated from the fixed foundation by 96 air springs and 80 dampers.
Chapter 2. Brief Review of the Existing Facilities

Specifications for NCREE

Dimensions 5x5 m
Mass 27 tons
Payload 50 tons
DOFs 6
Maximum force capacity (X, Y, Z axis) 220 kN, 600 kN, 600 kN
Stroke (X, Y, Z axis) +/- 250 mm, +/-100 mm +/- 100 mm
Peak velocity (X, Y, Z axis) 1 m/sec, 0.6 m/sec, 0.5 m/sec
Peak acceleration (X, Y, Z axis) 1 g, 3 g, 1 g
Maximum overturning moment 1500 kN m
Maximum yaw moment -
Yaw rotation -
Operation frequency 0-50 Hz
Table 6 - NCREE specifications

Figure 13 - NCREE - View of the laboratory during dynamic testing

Chapter 2. Brief Review of the Existing Facilities

2.6 UK - Bristol (EQUALS)

The Earthquake and Large Structures Laboratory (EQUALS) is part of the Bristol
Laboratories for Advanced Dynamics Engineering (BLADE) in the Faculty of Engineering at
the University of Bristol. It houses a 15 tons capacity, 6 DOFs earthquake shake table
surrounded by a strong floor and adjacent strong walls up to 15 m high.

The shake table consists of a stiff 3 by 3 m platform, weighing 3.8 tons. The platform can
accelerate horizontally up to 3.7 g with no payload and 1.6 g with a 10 tons payload.
Corresponding vertical accelerations are 5.6 g and 1.2 g respectively. Peak velocities are 1
m/s in all translational axes, with peak displacements of ± 0.15 m.

The shake table is accompanied by a set of 40 servo-hydraulic actuators that can be

configured to operate in conjunction with the shake table, strong floor and reaction walls,
providing a highly adaptable dynamic test facility that can be used for a variety of earthquake
and dynamic load tests.

Hydraulic power for the shake table is provided by a set of six shared, variable volume
hydraulic pumps, providing up to 900 l/min at a working pressure of 205 bar. The maximum
flow capacity can be increased to around 1200 l/min for up to 16 seconds at times of peak
demand with the addition of extra hydraulic accumulators.

A special feature of the EQUALS facility is its digital control system, with world leading
features, including a ‘hybrid test’ capability (also known as ‘dynamic sub-structuring’) in
which part of the structural system of interest can be emulated by a numerical model
embedded in the digital control system, while only a sub-component need be tested
physically. Extensive instrumentation is available, including 256 data acquisition channels.

The research based on the EQUALS shake table includes the response of cable-stayed
bridges, soil-structure interaction, the use of discrete damping elements in building structures,
base isolation systems, torsional response of buildings, masonry structures, steel and concrete
buildings, multiple-support excitation, travelling earthquake wave effects, non-linear self-
aligning structures, dams, reservoir intake towers, retaining walls and strengthening systems
with advanced composites.
Chapter 2. Brief Review of the Existing Facilities

2.7 PORTUGAL - Lisbon (LNEC-3G)

The Earthquake Engineering Research Centre (NESDE) at LNEC has a large 3D shake table
(LNEC-3G) located in a large testing hall with floor-to-ceiling height of 10 m, enabling
testing of tall structures.

Figure 14 - Arrangement of the three-axis table at LNEC [ECOEST. 1997]

LNEC-3G has three independent translational DOFs, with rotational ones minimised via a
torque tube system. Under the horizontal cranks, passive gas actuators may enable peak
velocities up to 0.7 m/s.

LNEC’s current 3D shake table was designed specifically for testing civil engineering
structures and components up to collapse or ultimate limit states. The steel platform measures
5.6 by 4.6 m and weighs 40 tons. It can carry a maximum payload of 40 tons. In terms of
capabilities, the system reaches a peak acceleration of 1.8 g (with zero payload), peak velocity
of 0.2 m/s and a stroke of ± 175 mm. The platform is driven by one 1000 kN longitudinal, two
300 kN lateral and one 300 kN vertical servo-hydraulic actuator situated on the centerlines of
the table. The testing frequency limit is 20 Hz.

LNEC-3G capacity in terms of payload (40 tons) allows testing of small real scale buildings
or larger buildings at smaller scales (bridge piers and 4-storey buildings have been tested at
1:3 scale).
Chapter 2. Brief Review of the Existing Facilities
Chapter 2. Brief Review of the Existing Facilities

2.8 FRANCE - Saclay (AZALEE)

The TAMARIS infrastructure and its main shake table AZALEE belong to CEA’s Seismic
Laboratory. The AZALEE shake table, with 100 tons allowable model mass, is the largest
shake table in Europe. To date, tests with masses up to 92 tons have been successfully
performed. The shake table has dimensions of 6 by 6 m and 6 DOFs allowing testing of
specimens under independent excitations of various types: sinusoidal, random, shock and
time-history with 0 to 100 Hz frequency ranges.

Maximum accelerations of 1 g and 2 g in the horizontal and vertical directions, respectively,

can be applied to specimens with the maximum payload of the table. The peak velocity of the
shake table is 1 m/s, peak displacements are ± 0.125 m and ± 0.1 m in the horizontal and
vertical directions, respectively.

The areas of research supported by the infrastructure cover a variety of experimental and
analytical RTD projects, both in the nuclear and non nuclear fields, for equipment, buildings
and soil-structure interaction; both new and existing structures are addressed. Assessment and
retrofitting of existing buildings and equipment are of special interest for the laboratory.

Figure 15 - AZALEE - View of the laboratory

Chapter 2. Brief Review of the Existing Facilities

2.9 ITALY - Pavia (TREES Lab)

The activities of the EUCENTRE benefit from one of the most advanced laboratories in the
world (TREES Lab – Laboratory for Training and Research in Earthquake Engineering and
Seismology). The testing facilities available at the EUCENTRE TREES Lab consist of a large
unidirectional high performance shake table, a reaction system composed of two L-shaped
reaction walls and a strong floor and an advanced high performance bearing tester system.

The main specifications of the experimental facilities are:

- Shake table. The SDOF system with dimension of 5.6 by 7.0 m in plan has maximum
rigid payload of 140 tons, peak displacement of ± 500 mm, peak velocity of 2.2 m/sec,
peak acceleration with maximum payload of 1.8 g, maximum force of 2100 kN and
maximum overturning moment capacity of 4000 kNm.
- Bearing Tester System. The system is composed by a 5 DOFs table with dimension of 1.7
by 2.0 m. Main performance characteristics are: force capacity of 2000 kN and 50000 kN
for the horizontal and the vertical direction respectively, displacement capacity of ± 600
mm and ± 75 mm for the horizontal and the vertical direction respectively and maximum
acceleration of ± 1.5 g.
- Hydraulic Power System. Propulsion for the actuators is provided by a mixed system
comprising 8 pumps for a total continuous flow of 1440 l/min and by 7 groups of
accumulators capable of increasing the total flow to 11000 l/min when running dynamic
tests on the shake table and up to 27000 l/min when used with the BTS.
- Data Acquisition System. It includes a 250-channel system based on 18 bit hardware and
an advanced wireless system based on 8 high definition digital cameras.

Access to the shake table will be given to projects focused on seismic risk reduction involving
dynamic studies on scaled or real scale structures, using concrete, masonry, steel or wooden
prototypes. Alternatively, tests can be performed on subassemblies or parts of structures
whose behavior can be conveniently and separately investigated from the whole structure.

Access to the Bearing Tester System will be given for cyclic static or dynamic seismic tests
on bearings and seismic isolation devices, performance assessment and prototype study and
development. User activities can be oriented to traditional devices (rubber bearings, dampers,
pots, etc.) or to innovative devices, such as those based on pendulum (FPS, friction pendulum
system with single and double curvature), magneto-rheological, etc.
Chapter 2. Brief Review of the Existing Facilities

Figure 16 - TREES Lab - View of the laboratory

Specifications for TREES Lab (shake table)

Dimensions 5.6x7 m
Mass 43 tons
Payload 140 tons
DOFs 1
Maximum force capacity 2100 kN
Stroke +/- 500 mm
Peak velocity 2.2 m/sec
Peak acceleration (bare table) +/- 6 g
Maximum overturning moment 4000 kN m
Maximum yaw moment 400 kNm
Yaw rotation -
Operation frequency 0-50 Hz
Table 7 - TREES Lab specifications
Chapter 2. Brief Review of the Existing Facilities

2.10 Comparisons
The comparison between the specifications of each of the selected facilities it’s important in
order to have a better understanding of what the specifications of the new laboratory should
be.

Laboratory Location Payload Peak acc. Peak vel. Stroke [mm]

LHPOST San Diego, USA 2000
[tons] 1.2
[g] [m/sec]1.8
[m/sec] +/-750
SEESL Buffalo, USA 50 1.15 1.25 +/-150
NEES @ Reno, USA 45 1 1.27 +/-300
E-DEFENSE Miki-city, JAPAN 1200 1 2 +/-1000
NCREE Taipei 50 3 1 +/- 250
EQUALS Bristol, UK 15 1.6 1 +/-150
LNEC-3G Lisbon, PORTUGAL 40 1.8 0.2 +/-175
AZALEE Saclay, FRANCE 100 1 1 +/-125
TREES Lab Pavia, ITALY 140 1.8 2.2 +/- 500
TREES Lab 2 Pavia, ITALY 120 2 2 +/- 250
Table 8 - Comparison of the performances for the considered shake table systems

The previous table shows that the new laboratory TREES Lab 2 has the second highest
performance in terms of both PGA and PGV and a considerable value for the payload. The
intentions for the design of it, more clearly described in the following chapter, were to have a
shake table system able to perform acceleration records with high demand in terms of PGA
and frequency content so to reproduce also the near fault effects on the response of the tested
specimens.
Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

3 INTRODUCTION OF THE NEW TREES LAB 2 AT

EUCENTRE
After a general review of the most efficient existing facilities and a comparison between them,
this chapter will describe all the initial considerations made about the new facility. In the
picture below it’s showed the area in which the new building will be constructed. This area
contains the actual laboratory TREES Lab, the building of Eucentre in which are located most
part of the offices and the Department of Structural Mechanics.

Figure 17 - Position of Eucentre (Pavia, Italy)

Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

3.1 Location
Since 2005, Eucentre Foundation performs research activities in the field of the reduction of
the seismic risk, by using high-tech equipment. The laboratory is equipped with one of the
most powerful shake table of 1 degree of freedom in Europe in terms of load and
displacement capacity and with a Bearing Tester of 5 DOFs for tests of bearing devices with a
vertical load capacity up to 50000 kN. Those devices are supported by a hydraulic system that
is able to provide up to 27000 l/min of oil at a pressure of 28 MPa. The realization of the new
laboratory next to the main building of Eucentre would allow improving the capacity of the
research centre with a new shake table, being able in this way to satisfy the requests of the
industries in the field of big scale multi-axial tests. It has also to be considered the importance
of the knowhow and the human resources already present at the Eucentre Foundation that
consists of eighty researchers and expert technicians in the field of the dynamic tests. This
represents a wealth of experience already available for the new laboratory.

3.2 Objectives and applications

Define in detail the characteristics of an advanced research laboratory in the field of
earthquake engineering means that there is the need to satisfy the requirements of modularity,
flexibility and complementarity considering the existing research centres. One of the most
important objectives in the design of the laboratory is to provide the possibility of perform
test of structure without limit in terms of capacity.

In order to do that it’s necessary:

- obtain high performance in terms of maximum load capacity, velocity, displacement,

acceleration and control system;
- be flexible in the possibility of building and moving the specimens of the test;
- create a strong integration between the new laboratory, the existing one and the laboratory
net already working around the world.

In order to satisfy the requests of the industry in terms of multi-axial tests, the fundamental
requirement of the project is the design of a shake table with 3 degrees of freedom that can
be extended up to 6 degrees of freedom. To do that there is the need to have an important
amount of oil in pressure with a sensible increment on the cost of the project (pumping
system). In order to solve this issue, the idea is to connect the shake table to the existing
hydraulic system by adding an accumulation system. This will lead to a savings of money for
the project itself.
Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

3.3 Requirements of the facility

Starting from the needs of the new laboratory those are the important points:

- the idea is to design a new facility in order to satisfy the increasing requests for big scale
tests;
- the dimension of the new shake table (7 by 7 m) represents an economical agreement for
the realization of big scale tests;
- the dimension of the laboratory is weighted on the kind of performed test, the necessary
safety areas for the workers and the structures and the necessity of moving specimens with
sensible dimension;
- the strategic position of the new laboratory is function of sharing the construction area
with the existing laboratory, where it’s possible to move full scale specimens, and also
sharing the hydraulic system;
- the pedestrian paths and the path for the vehicles are function of a perfect integration with
the existing structures.

Figure 18 - Prospectus of the laboratories with inner view

Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

3.4 Local needs and constraints

When a dynamic and pseudo-dynamic testing facility is designed, the main goal is to develop
a laboratory that provides the possibility of performing experimental tests on structures
without limitations in terms of capacity. For example, the specifications of a shaking table
should define a facility capable of simulating the mechanisms of collapse of structures under
strong earthquake motions.

To simulate real three-dimensional earthquake motion, the first requirement is the number of
allowed degrees of freedom (DOFs) of the shake table. They have at least to be three.
Secondly, the size of ground motion that can be reproduced is normally governed by its
velocity content. This is directly related to the oil flow rate that can be provided by the
pumping system and the servovalves.

Talking about the possible type of specimens, they mostly have a rectangular shape so that a
rectangular table area is required. The current worldwide design trend is to build facilities
whose velocity and displacement performance limits are emphasized more than acceleration.

The design of the table follows this trend of high velocity and large amplitude requirements in
order to reproduce the size of real earthquakes. Finally, the characteristics of this new testing
facility have been chosen so as to complement, rather than to compete, with existing
earthquake engineering testing facilities in Europe, and, in particular, in Italy.

3.5 General description of the new dynamic testing facility specifications

The project provides the construction of a new prefabricated structure next to the existing
laboratory with the same size. The new 7 by 7 m shake table together with the RC foundation
mass is positioned in the centre of the laboratory area. The displacements in the two
horizontal directions are guaranteed by 8 high performance actuators connected on two
sides of the steel structure of the shake table with an angle of 90 degrees between them. To
have high performances in terms of frequency there was the need of having 4 symmetrically
positioned actuators for each of the two horizontal directions, two for each side of the table.

The shake table is also simply supported by 4 vertical pistons (pads) above which it can slide
because of a thin layer of oil in pressure (in the case of an upgrade to 6 degrees of freedom,
the vertical pads have to be substituted by vertical actuators). In order to resist to the
overturning moment (consequence of a dynamic test), there is the need to provide 2 hold-
down actuators (working with gas) that are responsible to prevent the overthrow of the
facility.
Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

The capacity to accurately reproduce near-fault ground motion effects was considered
essential for the facility since, the seismic hazard of the built environment is controlled by
near-field ground motion at long return period hazard (e.g. 2% probability of exceedence in
50 years) for many sites. Near-field, fault normal, ground motion records with forward
directivity effects (Doppler effects) are characterized by a large velocity pulse, while near-
field, fault parallel, ground motion records are characterized by a fling step (i.e. large step
function in the ground displacement record). This is controlled by the peak table velocity
parameter. A peak table velocity of 2 m/s (both directions) was selected according to the
available near-fault seismological data and as a compromise between technical performance
and budgetary constraints. For the reproduction of far-field ground motions, a maximum
horizontal peak ground and peak table acceleration of 2 g (both directions) corresponding
to an upper bound of the vast majority of recorded ground motion records, was considered.
The significant frequency content of actual earthquake horizontal ground acceleration records
lies in the range between 0 to 15 Hz, while the significant frequency component of horizontal
ground velocity and displacement records lie in a lower frequency range than that of the
acceleration. A frequency bandwidth of 30 Hz for accurate reproduction of actual full-scale
ground acceleration records by the table was chosen.

Specific analyses have to be done in order to consider the needed performance for
reproducing the acceleration records selected to test the specimen. Several earthquakes will be
selected for the analyses with different acceleration, velocity and frequency contents.

With regard to the construction technology, also considering the actual shaking table at
TREES Lab, a steel welded platform was chosen. The platform consists of a closed steel box
structure constructed in a single piece. Several tables are made in steel or aluminium. The
table could be made from several aluminium sections bolted and glued together. The platform
surface could be an arrangement of aluminium plates or even magnesium plates. Since the
table has to survive experiments without deformations, the stiffness of steel made it preferable
compared to aluminium.

A very stiff table with a honeycomb-like network of stiffening diaphragms giving it

strength and bending stiffness was chosen. The internal steel sheets run in longitudinal and
transversal direction for the overall length and width of the platform. Horizontal internal
plates have also been added. The table becomes very stiff with high fundamental frequency
response, which is one of the design requirements for a straightforward control. There is, in
fact, the need of being far away from the common frequencies of the kind of specimens that
can be tested on it. The positions of the internal diaphragms and their spacing are also related
to the hydrostatic bearing dimensions and locations.
Chapter 3. Introduction of the New TREES LAB 2 at Eucentre

Considering the specification described before, the selection of these elements was made:

- MTS actuator model 244.51S with a force capacity of 1000 kN;

- MTS hold-down model STRUT 675;
- Pads with a diameter of 100 cm.

Talking about the performances, taking into account also the geometrical constraints, if the
facility works with 3 degrees of freedom, the facility itself is able to perform about 25 cm
displacement on the two horizontal directions and about 5 degrees of rotation about the
vertical axis. By moving 4 of the 8 actuators on one of the horizontal directions, there is the
possibility of using the shake table as a SDOF system. This layout can significantly increase
the performance of the facility in terms of frequency and force.

Specifications for TREES Lab 2

Dimensions 7x7 m
Mass 50 tons
Payload 120 tons
DOFs from 3 up to 6
Maximum force capacity 2000 kN
Stroke +/- 250 mm (both directions)
Peak velocity +/- 2 m/sec (both directions)
Peak acceleration 2 g (both directions)
Maximum overturning moment 5000 kNm (both directions)
Yaw rotation +/- 5 degrees
Operation frequency 0-30 Hz
Table 9 - General characteristics of the new shaking table at TREES Lab 2

Finally the project involves the placing of a reinforced concrete foundation mass of around
1200 mc with a weight of 2000 tons. On the surface of the foundation mass, modular pre-
stressed concrete blocks will be placed for the anchorage and the tight of the post-tensioning
bars. It’ll be possible in this way to monitor the reaction wall according to the needs of each
test.

All the specifications just described are going to be the starting point for the project of the
new laboratory and specifically of the new shake table. It’s important to clarify that more
analyses than the ones done for this work are needed in order to define clearly and in detail
the characteristics of the laboratory. At the specific moment of this work, all the analyses
made are meant to be guidelines for the future development of the project.
Chapter 4. Component of a Shake Table

4 COMPONENT OF THE SHAKE TABLE FACILITY

The principal components involved in the design of a shake table facility are actuators,
bearing pads, bold-downs, servovalves, accumulators, hydraulic pumps, piping system,
cooling system, controllers and associated sensors. The following paragraphs will describe
qualitatively the components referring also to the specifications for TREES Lab 2.

4.1 Actuators
The actuator is the device able to reproduce the input record chosen to test the specimens.
According to the way of acting of the device, the actuators can be divided in the 3 different
types as shown in the following figures.

Figure 20 - Double acting Figure 21 - Double acting

Figure 19 - Single acting
actuator - Single rod actuator - Double rod
actuator

The double acting actuator with double rod has more advantages comparing to the others. It
allows reducing distortion, having a symmetric behaviour at the same time and having an
upper lateral frequency. Mainly it allows having a better control with symmetric servovalves.
Chapter 4. Component of a Shake Table

Talking about the mechanics of the device, an actuator is composed by the following parts:

- a cylinder with hydrostatic bearings or special bearings with very low friction coefficient;
- a piston;
- two swivels (one at each extremities);
- one or several servovalves;
- a LVDT transducer in the piston;
- a small LVDT in the servovalves;
- a pressure sensor;
- load cells;
- hydraulic cushions to reduce the impact force when the piston arrives at full stroke.

The actuators then, have to be connected to the shake table. For mono-axial test and high
frequency excitation, the actuator can be linked rigidly to the table. In the majority of the
cases, for mono axial or multi axial shake table, the swivels have to be used to allow rotations
between the table and the foundation system. The swivels are located at each of the extremity
of the actuators. This configuration decreases the lateral vibration level applied to the actuator
and to the servovalves.

The choice for TREES Lab 2 has been made preferring more actuators with a smaller size
instead of a low number of bigger actuators. In this way the shake table system will have a
better performance in terms of frequency. The scheme of the shake table system consists in 8
horizontal actuators, 2 actuators for each side of the shake table so that 4 actuators are
working for each of the 2 horizontal directions.

4.1.1 Force
Given the size of the shake table, payload and the mass of each of the actuators, actuator
forces can be estimated. As previously written the payload is 120 tons, the shake table mass is
around 50 tons and the target acceleration is 2 g for both directions. Considering also the mass
of the 8 actuators (1 ton each), the 2 hold-downs (l ton each) and the 4 hanging bearings (1
ton each), the force needed for both directions is equal to:

F=ma=( 120+50+14 )∗2∗9.81=¿ 3610 kN

Chapter 4. Component of a Shake Table

4.1.2 Specifications of the actuator

According to the preliminary calculation in terms of force demand and considering that, for
one horizontal direction there are 4 actuators, the MTS actuator model 244.51S with a force
capacity of 1000 kN is able to to satisfy the force requirements.

Figure 22 - Cutaway view of MTS Series 244 hydraulic actuator

Chapter 4. Component of a Shake Table

1. Piston rod end. The piston rod has replaceable, hardened-steel insert that provides an
internal thread for mounting force transducers, grips, swivels, etc.
2. Porting. Large ports are designed to accept servovalve flow rates up to 90 gpm (340
l/min).
3. Piston rod. Series 244 Hydraulic Actuators use one-piece, double-ended piston rods. The
double-ended design improves performance by eliminating the compromises caused by
unequal tension and compression piston areas. The single-piece piston rod is heat-treated
alloy steel that is hard chrome plated, then ground and polished to a precision surface
finish using a machining process. This process improves wear resistance and extends seal
and bearing life. The gun-drilled rod permits convenient installation and accurate
alignment of a linear variable displacement transducer (LVDT).
4. Cushions. Upper and lower hydraulic cushions protect the actuator from the adverse
effects of accidental high velocity impacts between the piston rod and end cap.
5. Piston seal. A reinforced polytetraflouroethylene (PTFE) seal on the piston provides a
positive seal and reduces friction.
6. Piston rod seals. The piston rod seals are designed for long life, low friction and
exceptional performance in high-frequency, low-displacement applications. The piston
rod seals consist of a high-pressure seal Cutaway View of Series 244 Hydraulic Actuator
and a low-pressure/wiper seal in both the front end cap and the rear end cap. A small
amount of hydraulic fluid is allowed to flow past the high-pressure seal for continuous
bearing lubrication. Drainback ports return the hydraulic fluid passed by the high-pressure
seal back to the system drain line. The inner part of the low-pressure/wiper seal wipes
hydraulic fluid passed by the high-pressure seal from the piston rod and guides it to the
drainback ports. The outer part of the seal functions as a scraper ring to minimize external
contamination of the seals and bearings.

Figure 23 - Channel system

Chapter 4. Component of a Shake Table

4.1.3 Swivel rod end

A swivel rod end is a spherical bearing with a patented
adjustable preload mechanism to minimize backlash.
Proper adjustment of the bearing preload reduces impact
noise, minimizes friction, and improves the waveform
during load reversals encountered in dynamic
tension/compression testing. The swivel rod end uses a
connector stud and one set of spiral washers.

4.1.4 Spiral washer

Spiral washers fit over the studs at each connection and are
rotated in opposite directions (producing a camming effect)
to a position that places a constant preload on the stud.
They create a backlash free assembly that fully preloads the
connections between the actuator, swivel, and force
transducer.

4.1.5 Force transducers

Series 661 Force Transducers produce an electrical signal
proportional to the force exerted by the actuator. They
achieve at least 99% accuracy throughout the actuator force
range.

Figure 24 - Actuator series 244

4.1.6 LVDT
A linear variable differential transformer (LVDT) provides a piston rod displacement
feedback signal to the system control electronics. The LVDT is coaxially mounted within the
actuator piston rod and produces an analog signal with excellent linearity

4.1.7 Swivel bases

A swivel base provides the same features as the swivel rod end. For structural testing
applications, a Series 244 Hydraulic Actuator equipped with a swivel rod end and a swivel
base allows the actuator to pivot freely at both ends. This configuration reduces moment loads
on the test specimen and load cell. The swivel base design attaches to a closed housing
LVDT.
Chapter 4. Component of a Shake Table

The specifications for the Series 244 Hydraulic Actuator models are shown on the following
pages.

Table 10 - LVDT specifications

Chapter 4. Component of a Shake Table

Figure 25 - Characteristics of the selected actuator

Figure 26 - Characteristics of the actuator swivel

Chapter 4. Component of a Shake Table

Figure 27 - LVDT specifications

4.1.8 Typical actuator configurations

Series 244 Hydraulic Actuators are used every day in labs throughout the world for rigorous
testing of materials, structures, and components. Some of the most common configurations
are listed below:
Chapter 4. Component of a Shake Table

- Structural Actuator. The most common configuration for the Series 244 Actuator is as a
component in a structural testing system. By adding a swivel rod end, load cell, closed
housing LVDT, and swivel base to the basic actuator assembly, it’s possible to mount the
actuator in any position.
- Load Frame Actuator. The compact size, high side load tolerance, and high force capacity
of the Series 244 Actuators make them ideal for load frame applications.
- Pedestal Base Actuator. Although most applications profit from at least a small degree of
pivotal freedom, others require rigid actuator fixturing. Pedestal bases are available for all
actuator models to handle these situations. Vibration testing, seismic tables, and horizontal
cantilever installations are common applications for pedestal bases

4.1.9 Flow
For one actuator the flow is the sum of several flow components:

- the kinetic flow due to the displacement of the pistons, required to produce a desired
velocity given by the section area of the piston times the velocity of the reproduced signal;
- the flow due to the compressibility of the oil which is primarily related to volume of oil in
the actuator, frequency of operation, load on the actuator and fluid bulk modulus;
- the cross coupling flow due to the displacement of the actuators in the other axes; when
the table moves in one direction the actuators in the other directions must also move for
geometric reason;
- the flow due to the overturning moment (distance of the centre of gravity of the specimen
to the centre of gravity of the shake table in the vertical direction);
- the flow due to the yaw moment reaction (horizontal offset of the centre of gravity);

Among them the main contribution comes from the kinetic flow directly connected with the
input records selected to be reproduced by the shake table system.

4.2 Bearings
A number of 4 bearing pads will be fixed below the shake table in order to allow the sliding of
the system during the tests. There are two options for the position of the bearing pads: the first
is to fix them to the table so that the table doesn’t change the boundary conditions during the
motion and the second is to fix them to the concrete sliding foundation support. The first
option has been selected in order to simplify the boundary conditions of the finite element
model representing the shake table.

According to observations and comparisons with already existing facilities having same
performances, pads with 100 cm diameter are initially considered. Finally, the possibility of
substituting the bearing pads with vertical actuators allows increasing the DOFs up to 6,
considering that the standard configuration of the system has 3 DOFs.
Chapter 4. Component of a Shake Table

4.3 Hold-Downs
In order to keep the table horizontally and overcome the presence of the overturning moment,
2 hold-downs are fixed vertically below the table. The force that the 2 hold-downs have to
apply has to be calculated in order to avoid that the bearing pads attached to the shake table
would lose the contact with the sliding surface. With the payload centre of gravity above the
shake table surface, adding then the length corresponding to the difference between the shake
table centre of gravity and the horizontal actuator plane, the 2 g acceleration in one of the two
horizontal directions generates the overturning moment. The height of the laboratory allows
test of specimens with a height up to 12 m. Considering the centre of gravity of the specimen
(max 120 tons) 6 m above the shake table surface and the centre of gravity of the table around
0.5 m below the surface itself, the overturning moment can be initially estimated:

M =( 120∗2∗9.81 )∗( 6+ 0.5 )=¿15305 kNm

Figure 28 - Parts of the selected Hold-Down

Figure 30 - Characteristics of the

Figure 29 - Characteristics of the selected Hold-Downs
Hold-Down swivel
Chapter 4. Component of a Shake Table

4.4 Servovalves
The size and number of servovalves is a function of peak velocity, peak acceleration,
frequency bandwidth of interest and correlation between axes (in the case of X and Y since all
eight horizontal actuators are used in these two directions).

The analyses will indicate a certain value for the system performance in terms of kinetic
flows. If, for example, 2 valves are used, then the flow requirements will be the half of the
kinetic flow per valve. The calculations of the kinematic servovalves requirements, given the
system specifications, are done with the following procedures:

Apiston = 487.7 cm2 Vpeak = 2 m/sec with 1 liter=1000 cm 3

cm3 c m3 −3 liter liter sec

Q= A piston ¿ V peak =487.7∗( 2∗100 )=97540 =97540 ∗10 =97.54 =97.54 ∗60 =5852
sec sec sec sec min

A number of 4 servovalves MTS 256.40 with a flow capacity of 1500 l/min each are selected
to satisfy the flow requirements.

Figure 31 - Cutaway view of a 256 servovalve

Chapter 4. Component of a Shake Table

The main features of the selected servovalves are:

- flow rating of 1500 l/min;

- electronic spool position feedback to the control loop to allow precise spool position
control for enhanced dynamic response;
- very low current threshold that permits high servovalve control resolution;
- matched, ground metering edges to minimize crossover distortion;
- high spool-driving force and hardened tool steel metering edges to reduce contamination
sensitivity.

Figure 32 - Series 256 servovalve

This servovalve is an electro-hydraulic three stage servovalve consisting of a high-flow, four-

way spool valve and a smaller two stage pilot valve. As shown, the first stage of the pilot
valve is the torque motor stage. This stage consists of coils, an upper and lower pole piece, an
armature and two magnets. Its purpose is to convert an electrical input signal (servovalve
control signal) into physical movement of the armature. The second stage of the pilot valve
consists of a spool valve which controls hydraulic fluid flow to the third stage, and a feedback
spring which provides mechanical linkage between the armature/flapper and the spool.
Chapter 4. Component of a Shake Table

The third stage is similar to the second stage except that it has a much larger spool valve and
its position is monitored by electrical feedback produced by a linear variable differential
transformer (LVDT). The LVDT provides an output signal to the valve controller that is
proportional to the spool position. This signal is compared to the desired position (system dc
error signal) in the valve controller. Sensing any difference between the system dc error and
the LVDT signal causes the valve controller to alter the servovalve control signal current to
the torque motor accordingly. Third (main) stage flow is applied to the actuator through the
control ports.

Figure 33 - Typical closed-coop servo-hydraulic testing System

Table 11 - Static servovalve performance

Chapter 4. Component of a Shake Table

Table 12 - Dynamic servovalve performance

Figure 34 - Servovalve flow-frequency Figure 35 - Servovalve flow-frequency

performance curve 256.40A performance curve 256.40B

Chapter 4. Component of a Shake Table

Figure 36 - Servovalve dimensions

Chapter 4. Component of a Shake Table

Table 13 - Servovalve dimensions table

Chapter 4. Component of a Shake Table

4.5 Oil column resonance

The oil column resonances are the predominant resonance of the system and typically lie well
within the operating frequency range of the system. The principal effect they have on system
sizing is that they mark the point on the performance curve at which the compressibility flow
requirements start to have a significant role as opposed to the kinematic flow requirements
that dominate at lower frequencies. Oil column resonances also have a crucial impact on
system fidelity.

4.6 Lateral resonance effects (bow string frequency)

The effects of the lateral resonance are different considering single ended actuators or double
ended actuators. The main parameters acting on the bow string frequency are:

- the length of the actuator;

- the section of the piston;
- the section of the cylinder.

The lateral resonance issue is a serious factor that has plagued several earthquake simulators.
It has resulted in delays during acceptance lasting several months or more. It is a problem that
typically afflicts suppliers as they try to supply multiple or 6 DOFs earthquake simulator’s for
the first time. The solutions for this problem have been successfully implemented on 10 large-
scale multi degree of freedom systems since 1990.

All actuators and passive restraining links have a natural lateral resonance often referred to as
the 'bow string' resonance due to its modal shape. In designing systems a compromise is
struck between actuator stroke and high frequency performance.

The effect of this resonance is to create a 'hole' in the transfer function of the system. The hole
is larger for actuators with lower resonant frequencies due to lower effective damping within
the actuator structure. The magnitude of this resonance can impose severe stress levels on the
actuator structure resulting in the potential for catastrophic failure.
Chapter 4. Component of a Shake Table

4.7 Accumulators
The actuators are alimented by means of pumps which give oil (continuous flow) with a given
pressure and temperature. To increase the performances of the system and to take into account
the fast variation of pressure and flow during dynamic motions, there is the need of having a
sufficient accumulator capacity. The hydraulic system works as follow. When the flow
through the actuator decrease, the pressure increases and the oil fills up the accumulator
(input). Instead, when the flow is over the continuous flow, the pressure decrease and the
accumulator discharges the oil through the actuator.

The objectives of the accumulators are:

- to increase the performances of the actuator (distortion and velocity);

- to decrease hydraulic shocks in the piping.

Some others accumulators can be placed directly on the hydraulic line to increase the power
or the global flow of the testing facility. The accumulators are pressurized when the pumps
starts and they give flow for a short time when the testing facility needs oil.

4.7.1 Bladder accumulators

A bladder accumulator consists of a pressure vessel with an internal elastomeric bladder with
pressurized nitrogen on one side and hydraulic fluid on the other side (system side). The
accumulator is charged with nitrogen through a valve installed at the top. The accumulator
will be pre-charged to nominal pressure when the pumps are not operating. When nominal
hydraulic system pressure is applied the bag will be compressed to its fully compressed state.

When the bag is fully compressed, the nitrogen pressure and the hydraulic pressure are equal.
As system pressure drops the bag expands, forcing fluid from the accumulator into the
system. When the bag expands then the pressure in the bag will decrease. The bag will
continue to expand until the bag pressure equals the hydraulic pressure (which will be lower
than nominal system pressure) or the bag fills the entire accumulator volume (an undesirable
situation). A poppet valve keeps the bag in accumulator from being pulled into the
downstream tubing should the bag over-expand.

If the bag was pulled into the downstream tubing, the accumulator would never recharge and
normal flow from the pump would be constricted. The maximum flow rate of the accumulator
is controlled by the opening area (orifice) and the pressure difference across the opening. The
main advantages of a bladder accumulator are fast acting, no hysteresis, not susceptible to
contamination and consistent behavior under similar conditions. Accumulators are easy to
charge with the right equipment. Because there is no piston mass, the speed of the bladder
accumulator is governed by the gas, which reacts very fast to changes in hydraulic system
pressure.
Chapter 4. Component of a Shake Table

The main limitation of bladder accumulators is the compression ratio (maximum system
pressure to pre-charge pressure) which is limited to approximately 4 to 1. Hence gas
accumulators will be larger than other accumulators for the same flow requirements.

The pre-charge pressure is typically set to approximately 80% of the minimum desired
hydraulic system pressure.

Figure 37 - Bladder accumulator schematic

Figure 38 - Bladder accumulator

Chapter 4. Component of a Shake Table

4.7.2 Piston accumulators

A gas piston accumulator has a piston which slides against the accumulator housing on seals.
On one side of the piston is nitrogen and on the other side is the hydraulic fluid and
connection to the system. A fill port allows pressurization of the nitrogen. A gas piston
accumulator will not respond to transient pressures as fast as a bladder accumulator due to the
mass of the piston (frequency characteristics depend on piston mass and spring characteristics
of the nitrogen). However, a piston accumulator will have better damping due to hydraulic
leakage (viscous damping) and friction between the piston and housing (coulomb friction &
seal friction). Piston accumulators may also be more prone to leakage than other types of
accumulators due to the seals. Piston accumulators will generally provide higher flow rates
than gas accumulators for equal accumulator volumes. This is because piston accumulators
can accommodate higher pressure ratios (maximum system pressure to pre-charge pressure)
than gas accumulators, up to 10 to 1, compared with bladder accumulator ratios of 4 to 1. The
disadvantages of piston accumulators are that they are more susceptible to fluid
contamination, have a lower response time than bladder (unless the piston accumulator is at a
very high pressure) and will have hysteresis from the seal friction. The pre-charge for a gas
piston accumulator is typically set to around 90% of minimum desired hydraulic system
pressure.

Figure 39 - Open view of a floating piston accumulator

Chapter 4. Component of a Shake Table

Each type of separated, hydro pneumatic accumulator has advantages, but bladder designs are
generally considered the most versatile. For shock and pulsation, for example, bladder models
are ideal. Piston units are not recommended because they are too slow to react to shock
waves. For emergency service, though, many users prefer piston accumulators. However, the
seals in piston types are not designed to hold pressure indefinitely without being cycled. The
system can leak and fail with no obvious indication to the outside. For piston and bladder
accumulators vertical mounting is the preferred orientation, with the gas side up.

Bladder designs predominate between one and 15 gallons, but are available to about 125
gallon capacity. The largest piston designs are typically about 100 gallon capacity.

Most accumulators used within industry are limited to an operating pressure of 3000 psi.
Accumulators are available which operate at higher pressures. In general, hydraulic
accumulators are pre-charged one half of the maximum operating fluid pressure; this is
adequate for most applications. For a system operating at 3000 psi, a properly rated
accumulator should be pre-charged (nitrogen is typically used) to 1500 psi.

Accumulators are typically rated by their manufacturer at gas volume when all fluid has been
expelled. The amount of fluid which can be stored within the accumulator is normally one
half of the gas volume.

Accumulators are selected based on the fluid pressure and volume requirements of the system
which they are to be installed into. The accumulator is sized such that the system fluid
pressure will not fall below a value resulting in degraded system performance.

As fluid enters the accumulator, the gas charge (normally nitrogen) is compressed. As the
fluid gas is compressed, the temperature will rise (Charles Law). Unless the gas is allowed to
cool to ambient temperature, elevated gas temperature will result in the volume of the fluid
entering the accumulator will be less than the calculated amount, resulting in the fluid
discharge to be less as well. Compensating for extra capacity may be facilitated by increasing
the require accumulator volume, 5% is a common value.

There are 2 types of accumulator systems: conventional and blow down.

4.7.3 Conventional accumulator system

Conventional accumulator systems are commonly used in many testing facilities by the world.
The accumulators can be charged to system pressure (210 bar) or higher (280 bar). The
pressure drops during operation but it is not usually a problem for most earthquakes testing.
For that solution, during the discharge of the accumulators to actuators, the pressure and the
performances decrease. The system characteristics are not constant and change during test,
and the transfer function of the table also.
Chapter 4. Component of a Shake Table

4.7.4 Blow down accumulator system

The accumulators are charged at much higher pressure (350 bar), and the system is always at
the standard pressure 210 bar. During tests, there are minimal pressure variations, typically
less than 10%. The system characteristics remain constant.

Figure 40 - Blow down accumulator bank system - MOC

4.7.5 Pumps versus accumulator blow down as a function of time

The hydraulic systems consist of pumps and accumulators. Pumps are designed to allow:

- a start-up in safety (the pilot pressure line must be always in pressure);

- tests at low velocity for sine sweep test and random excitation.

Accumulator racks are designed to allow seismic tests with a strong part of the excitation
during several seconds.

4.7.6 Horizontal half-sine shock on 120 tons specimen

The total dynamic force for one horizontal direction is 3610 kN. With this force the maximum
acceleration that can be generated on a 120 tons payload and 50 tons table is:

1000∗4 actuators
Acceleration= = 2.4 g
120+50 ton
Chapter 4. Component of a Shake Table

In reality the maximum acceleration will be lower due to added mass fixed to the table (a part
of the mass of the actuators must be added to the mass of the table).
Chapter 4. Component of a Shake Table

4.8 Hydraulic pumps

The best solution is to have a minimum of pumps to start the testing facility, to perform
random excitation and sine sweep at low level and to add additional accumulators to have the
full performance during a short time in good agreement with seismic tests requirement (2 m/s,
2 g simultaneously in the 2 directions).

Different equipments can be used for the oil circuit:

- diesel pumps;
- electric pumps;
- accumulators.

4.8.1 Diesel pumps

This type of equipment has been used for E-Defense in Japan. This equipment is very
expensive but the operating costs are low. The characteristics are the following:

- it is possible to increase the flow 30% (decrease pressure) by changing speed;

- there are noise issues and environmental issues which need to be investigated since each
engine can exceed 100 dB;
- the startup is faster since all diesel units can be started at same time;
- fuel consumption at low load and low rpm is low (< 10%);
- fuel consumption at full power is very high (several thousand liters/hour) which induce
especial requirements and issues for the storage and delivery of the fuel;
- diesel pumps require a maximum maintenance;
- diesel pumps take a large area.
Chapter 4. Component of a Shake Table

4.8.2 Electric pumps

Generally, all hydraulic testing facilities use electric pumps. The electric pumps need electric
substation, which increase costs, the power consumption at low flows (variable flow pumps)
is quite high (50%) and this choice limits the maximum size.

The main advantages of the electric pumps are:

- electric pumps are cleaner;

- the maintenance is minimal;
- the service pressure can be variable.

The principal parts of an electric hydraulic power system are:

- a motor;
- a pump;
- a pressure regulator;
- a tank;
- a filtration system;
- an oil cooler.

There are several types of pump on the market:

- gear pumps;
- vane pumps;
- radial piston pumps;
- axial-piston swash-plate pump.

The gear pumps are contained within a close-tolerance housing that separates the discharge
port from the intake port. An external shaft is connected to the driven gear, whereas the other
gear is supported by an internal shaft and bearing. To produce flow with a gear pump, fluid is
carried around the outside of each gear within each tooth gap from the intake side of the pump
to the discharge side of the pump. The gear pump exhibits advantages over other pumps in
two mains areas: the gear pump is inexpensive to buy, and there are only two moving parts in
the gear pump, which tends to make this design very reliable. Disadvantages of the gear pump
are as follows: the gear pump typically operates at a medium pressure level (up to 21 MPa)
efficiencies tend to be low (between 80 and 90%), gear pump tend be fairly loud (70 - 80 dB)
and gear pumps are fixed-displacement units, which means that the only control variable is
the shaft speed of the driven gear.
Chapter 4. Component of a Shake Table

The vane pump consists of an internal rotor, housing,

and spring loaded elements that are called vanes. The
vanes are contained within the rotor, and the rotor is
offset from the centre of the housing by the eccentricity
distance e. As the rotor turns relative to the fixed
housing, the vanes are thrown outwardly by their own
inertia. The springs also are used to assist in pushing the
vanes against the housing, and the discharge pressure is
sometimes supplied to the vanes as well for
accomplishing the same purpose. As the vanes pass over
the intake port on the housing, the space between the
vanes and the rotor and the housing increases, thereby
drawing fluid into the pump. As the vanes pass over the
discharge port on the housing, the space between the
vanes and the rotor and the housing decreases, thereby
pushing fluid out of the pump. Advantages and
disadvantages of the type of pump are similar to the gear Figure 41 - Cross sectional view of a
pump. vane pump

The radial-piston pump

consists of pistons that are
oriented in a radial fashion
within the cylinder block.
The intake and discharge
ports are located at the
centre of the cylinder
block. This pump has
features in common with
the vane pump in that it
depends on an eccentricity
offset shown by the
difference in the rotor
Figure 42 - Cross sectional view of a radial-piston pump centerline and the cylinder
block centerline. As the
rotor turns, the pistons follow the profile of the rotor, thereby drawing fluid into the intake
side of the pump and pushing fluid out of the discharge side of the pump. Advantages of the
radial piston pump are as follows: radial pumps typically exhibit high efficiencies (58 % – 98
%), they are capable of operating at high pressures (up to 42 MPa) and they exhibit a low
starting torque. Disadvantages of the radial-piston pump are as follows: it is a fairly expensive
pump to purchase and it is a fixed-displacement device, which means that the flow rate must
be varied by varying the shaft speed.
Chapter 4. Component of a Shake Table

The axial-piston swash-plate pump

consists of several pistons within a
common cylindrical block. The pistons
are nested on a circular array within
the cylinder block at equal intervals
about the shaft centerline. The cylinder
block is held tightly against a port
plate using the compressed force of the
cylinder block spring. A thin film of
oil separates the port plate from the
cylinder block, which under normal
operating conditions forms a
hydrodynamic bearing between the
Figure 43 - Cross sectional view of an axial piston
cylinder block and the port plate. The
cylinder block is connected to the swash plate pump
shaft through a set of splines that run
parallel to the shaft. A ball-and -socket joint connects the base of each piston to a slipper. The
swash-plate angle generally may vary with time to create a variable-displacement of the
pump. While the port plate is held in a fixed position against the head, the shaft and cylinder
block rotate about the shaft centerline at a constant angular speed. The shaft and the cylinder
block are connected through a set of splines that run parallel to the shaft. During this
rotational motion, each piston periodically passes over the discharge and intake ports.
Furthermore, because the slippers are held against the inclined plane of the swash plate, the
pistons also undergo an oscillatory displacement in and out of the cylinder block, and fluid is
drawn into the piston chamber. As the pistons pass over the discharge port, each piston
advances into the cylinder block, and fluid is pushed out of the piston chamber. This potion
repeats itself for each revolution.

Swash-plate axial-piston pumps are the most commonly used pumps in hydraulic control
system. Advantages of using these pumps are as follows: pumps exhibit high operating
efficiencies, are capable of operating a high pressure (up to 42 MPa) and swash-plate-axial-
piston pumps lend themselves nicely to variable-displacement control, which means that the
discharge flow from these pumps may be controlled by adjusting the swash-plate angle
independent of the pump shaft speed. Disadvantages of using these pumps are as follows:
swash-plate pumps are fairly expensive to purchase, and their construction is more complex
than the other pumps which leads to a lower degree of reliability.

One of the main problems due to the pumps is the noise (vibration) induce on the table.
Chapter 4. Component of a Shake Table

4.9 Cooling system

A cooling system is required to maintain constant the oil temperature at about 45 to 50 °C.

There are two kind of cooling system:

- air cooling system;

- water cooling system.

The air cooling system requires space, a system of forced inhalation and also generates an
important noise stress.

The feasibility bases itself on the hypothesis of departure of the differences between the air
temperature and that of the oil and the maximum return flow of oil. In these conditions it will
be necessary to plan a battery of coolers occupying a large volume.

The water cooling system, instead, may cool the circulating water at 4 or 5 ° C, which would
reduce water consumption and/or size of the cooler.
Chapter 5. Layout of the Shake Table System

5 LAYOUT OF THE SHAKE TABLE SYSTEM

For this project, the parameter governing the choice in the scheme of the horizontal actuators
is the operational frequency. In order to be able to reproduce an acceleration record with high
frequency demand (near fault events), there is the need of having the actuators acting along all
the shake table sides, aligned between them for each of the 2 horizontal DOFs. Then, for a
better control of the rotational DOF (about the vertical axis), the actuators should be
positioned at the edges of each of the shake table side.

The geometry of the foundation mass is designed in order to allow a different layout for the
system. All the 8 actuators can be set in order to act along one horizontal direction (SDOF
system), 4 on one side and 4 on the opposite one. This possible scheme will increase the force
capacity of the system for some specific tests.

It has to be pointed out that the final scheme of the facility, at least at the time of this work,
comes from few considerations made on the first draft of the project. Before going through to
the final scheme, a short description of the first draft and the reasons why it has been changed
are made.
Chapter 5. Layout of the Shake Table System

5.1 First draft

The first layout scheme of the project considered a 7 by 7 m shake table system with 4
horizontal actuators attached to the table web. In the following figure the system is shown.

Figure 44 - TREES Lab 2 - First draft layout scheme

The idea below this scheme was initially to have the possibility of saving space around the
shake table by connecting inclined actuator to the table web.

Figure 45 - TREES Lab 2 - First draft layout scheme

It has to be mentioned also that the system was since from the beginning a 3 DOFs system,
with longitudinal and transversal displacements and rotation about the vertical axis.
Chapter 5. Layout of the Shake Table System

5.2 Second draft

Talking about the second draft, only the position and the number of the actuators has changed.
As previously mentioned, the reason of this change lies in the need to perform tests with high
frequency demand.

Specifications First draft Second draft

Dimensions [m] 7x7 7x7
DOF 3 up to 6 3 up to 6
Mass [ton] 53 53
Longitudinal displacement [cm] +/- 100 +/- 25
Transversal displacement [cm] +/- 50 +/- 25
Yaw rotation [degree] +/- 10° +/- 5°
Longitudinal force capacity [kN] 2000 2000
Transversal force capacity [kN] 5000 2000
Maximum payload [ton] 120 120
Acceleration capacity [g] 2 2
Velocity capacity [m/sec] 2 2
Overturning moment [kNm] 5000 5000
Operational frequency [Hz] 0-15 0-30
Table 14 - Comparison between first and second draft

The following figures describe the characteristics and the geometry of the system by
considering the table first in its initial position and then during the maximum allowable
displacement along the Y direction (+/- 25 cm), the maximum allowable displacement along
X and Y direction simultaneously and the maximum allowable (5 degrees) rotation about the
vertical axis. Finally, the uniaxial configuration is presented.
Chapter 5. Layout of the Shake Table System

Figure 46 - Shake table layout - Position 0 of the actuators

Figure 47 - Shake table layout - Displacement along one of the two horizontal directions
Chapter 5. Layout of the Shake Table System

Figure 48 - Shake table layout - Combined displacement on the horizontal plane

Figure 49 - Shake table layout - Rotation about the vertical axis

Chapter 5. Layout of the Shake Table System

Figure 50 - Shake table layout - Position 0 in the configuration at 1 DOF

Figure 51 - Shake table layout - Displacement along the horizontal direction for the SDOF system
configuration
Chapter 6. Computation of the System

6 COMPUTATION OF THE SYSTEM

The performance specifications of a powerful dynamic test apparatus like a shaking table are
driven by the aim of creating a facility which will be able to simulate the responses of real full
scale structures under real strong earthquake motions. For these purposes, several
specifications are required.

6.1 Description of the numerical evaluations

A specific MATLAB program has been created on order to evaluate some of the specifications
needed for the design of the facility. A brief description of the implemented calculations is
done in this paragraph.

The sheet is divided in two parts, input data and calculations. The data required for the
calculations are summarized below:

- input acceleration record for X direction;

- input acceleration record for Y direction;
- input for the rotational DOF (about the vertical axis);
- time step of each of the three input records;
- scaling factor for each of the two acceleration records;
- time length of a zero vector to add at the beginning and at the end of each of the 3 records;
- maximum actuator extension and shortening for X and Y direction;
- initial length of each of the 8 horizontal actuators L0actuator (at the zero initial condition);
- initial inclination ϑ0actuator of each of the 8 horizontal actuators respect to the local vertical
axis of the actuators;
- angle ϑcentroid between the line connecting centroid of the shaking table (on the horizontal
plane) with hinge of the actuators (shaking table side) and the line passing parallel to the
side on which the actuators are attached;
- distance between each of the two actuator acting on same side of the shaking table,
perpendicular to the actuator movement for X and Y direction. The position of the two
actuators on the same side of the shaking table have to be symmetric respect to the
centerline of the shaking table;
- area of the piston for each of the 8 horizontal actuators;
- mass of each of the 8 horizontal actuators;
- mass of the specimen;
Chapter 6. Computation of the System

- weight per unit volume of the material used for the shake table skeleton;
- thickness of horizontal, transversal and longitudinal plates;
- area of each of the horizontal, transversal and longitudinal layers (plate alignments);
- distance between layer centroid and reference global system for each of the plate
alignments;
- continuous flow rate given by the pumping system.

Figure 52 - Zero initial condition of the adopted shake table system configuration

The input acceleration records are selected directly from the Pacific Earthquake Engineering
Research Center NGA Database. The PEER NGA (http://peer.berkeley.edu/nga/) database is
an update and extension to the PEER Strong Motion Database, first published on the web site
in 1999. The NGA database includes a larger set of records, more extensive meta-data, and
some corrections to information in the original database.

The created program is meant to be a quite flexible tool for different shake table geometry and
actuator layouts. It allows also the application of the baseline correction of the records
(subtract mean or best-fit line and from time series) and it’s possible to choose independently
which one of the three input records has to be reproduce by the shake table system.
Chapter 6. Computation of the System

In order to obtain the exact displacement that each of the actuator have to reproduce according
to their initial positions, all the three possible input records have to be combined together in
terms of displacement. This enables to account, in the evaluation of the total flow rate, for the
so called cross coupling flow. This contribution rises when the table moves only in one
direction and the actuators in the other directions must also move for geometric reasons,
affecting in this way the total flow rate and the total flow volume requirements.

The first step of the program is the evaluation of the velocity and displacement records by
integrating the two selected acceleration records. Baseline correction and scaling factors can
be chosen as option in the calculation.

At this point, the displacement records, for X and Y direction, and the rotation record are
combined together to obtain the displacement that each of the 8 horizontal actuators have to
reproduce. At each instant of the input records, the possible elongation/shortening of the
actuator piston according to the resultant of the 3 displacement components and the angle
between the actual actuator position and the initial zero position, is calculated. The maximum
displacement for each of the actuator is then checked with the maximum allowable
displacement due to the maximum stroke value.

The new actuator displacement records are now used to evaluate velocity and acceleration
records for the actuators by using the finite difference method. Finally it’s possible to evaluate
the modification of the initial input records due to the interaction between the initial signals
and the geometry of the shake table system.

As previously written, this first part of the sheet represents the input for the second part of it,
where the following quantities are evaluated afterwards:

- shake table mass;

- coordinates of the centre of gravity;
- force connected to the moving mass at each instant;
- flow rate for each of the actuators;
- total flow rate and total flow volume.

6.1.1 Displacement actuator demand

The procedure to get the displacement actuator demand is based on few steps. The
acceleration records are integrated twice to get the two displacement records and those are
then combined with the allowed yaw rotation (rotational input). Now, considering the actual
layout of the shake table system in terms of geometry, initial position of the actuators and
characteristics of the selected actuators, it’s possible to evaluate the local displacement
demand for each of the horizontal actuator.

This procedure is performed with the use of geometric parameters coming from the relative
position of the centre of the shake table with the position of the actuator swivel, the distance
between the two actuators on each of the table side, the initial inclination of the actuators and
the length of them. All these quantities involved are shown in the following figure
Chapter 6. Computation of the System

Figure 53 - Global reference system for the calculation of the actuator displacement

The calculation of the side lengths (ΔR, ΔX and ΔY) of the “reference” triangle is the starting
point and the key of all the calculations.

The method for the evaluation of the actuator displacement, based on the geometry of the
system, is shown for a possible horizontal displacement and rotation of the table. Finally a
figure shows the scheme for the combination of the 3 DOFs.
Chapter 6. Computation of the System

If only displacements input are considered for the shake table, the evaluation of the actuator
displacement is based on the fact that the hinge that fixes the actuator to the table follows
exactly the displacement of the centroid of the table. In the specific case of the figure, a
displacement of 20 cm along the X direction and a displacement of 25 cm along the Y
direction are provided to the table by the actuators. The calculations are showed for simplicity
for the case of actuator Y1.

Figure 54 - Displacement of the shake table

Chapter 6. Computation of the System

When a yaw rotation is applied to the shake table system, the angle ϑtriangle is required. This
angle belongs to the smaller “reference” triangle that comes from the new line connecting the
shake table centroid and the new position of the actuator swivel. In the specific case of the
figure, the angle ϑcentroid = 57.41° and the yaw rotation applied is equal to 20°. So, considering
the specific case of actuator Y1:

ϑtriangle = ϑcentroid + yaw rotation = 77.41°

Figure 55 - Rotation of the shake table

Chapter 6. Computation of the System

Finally, the 3 DOFs are combined together for the system. All the dimensions required for the
procedure are showed as follow.

Figure 56 - Combination of displacement and rotation for the shake table system
Chapter 6. Computation of the System

Displacement actuator Y1 – Y2

Input: displacement X direction, displacement Y direction.

∆X actuator
=actuator displ . component ∈X direction=displ .record X direction

∆Y actuator
=actuator displ . component ∈Y direction=displ .record Y direction

∆X ∗sin ( ϑ0
ϑactuator =−atan
[ L0
actuator
−L 0
∗cos ( ϑ 0
actuator
actuator

actuator
actuator

) +∆Y
)
actuator
]
−ϑ 0 actuator

L0 ∗cos ( ϑ0 ) + ∆Y
∆ actuator =
[ actuator

cos ( ϑ actuator +ϑ 0 )
actuator

actuator
actuator

] −L0 actuator

Input: displacement X direction, displacement Y direction, yaw rotation.

ϑtriangle = ϑ centroid+ yaw rotation→ for actuator Y 1

{
ϑcentroid − yaw rotation → for actuator Y 2

−[ ∆R∗cos ( ϑtriangle ) −∆ X ] + displacement X direction → for actuator Y 1

∆X actuator
=
{[ ∆R∗cos ( ϑ triangle ) −∆ X ] +displacement X direction→ for actuator Y 2

∆Y actuator
=−[ ∆ R∗sin ( ϑtriangle )−∆Y ] +displacement Y direction

∆X ∗sin ( ϑ0
ϑactuator =−atan
[ L0
actuator
−L 0
∗cos ( ϑ 0
actuator
actuator

actuator
) +∆Y
actuator

actuator
)
] −ϑ 0 actuator

L0 ∗cos ( ϑ0 ) + ∆Y
∆ actuator =
[ actuator

cos ( ϑ actuator +ϑ 0 )
actuator

actuator
actuator

] −L0 actuator
Chapter 6. Computation of the System

Figure 57 - Displacement for actuator Y1

Figure 58 - Displacement for actuator Y2

Chapter 6. Computation of the System

Displacement actuator Y3 – Y4

Input: displacement X direction, displacement Y direction.

∆X actuator
=actuator displ . component ∈X direction=displ .record X direction

∆Y actuator
=actuator displ . component ∈Y direction=displ .record Y direction

∆X + L0 ∗sin ( ϑ 0
ϑactuator =atan
[ L0
actuator

actuator
∗cos ( ϑ0
actuator

actuator
actuator

) −∆Y
)
actuator
] −ϑ0 actuator

L0 ∗cos ( ϑ0
∆ actuator =
[ actuator
) −∆Y
cos ( ϑ actuator + ϑ0 )
actuator

actuator
actuator

] −L0 actuator

Input: displacement X direction, displacement Y direction, yaw rotation.

ϑtriangle = ϑcentroid − yaw rotation → for actuator Y 3

{
ϑcentroid + yaw rotation → for actuator Y 4

−[ ∆R∗cos ( ϑtriangle ) −∆ X ] + displacement X direction → for actuator Y 3

∆X actuator
=
{[ ∆ R∗cos ( ϑtriangle ) −∆ X ] +displacement X direction → for actuator Y 4

∆Y actuator
= [ ∆ R∗sin ( ϑtriangle )−∆Y ]+ displacement Y direction

∆X + L0 ∗sin ( ϑ 0
ϑactuator =atan
[ L0
actuator

∗cos ( ϑ0
actuator
actuator

actuator
) −∆Y
actuator
)
actuator
] −ϑ0 actuator

L0 ∗cos ( ϑ0
∆ actuator =
[ actuator
) −∆Y
cos ( ϑ actuator + ϑ0 )
actuator

actuator
actuator

] −L0 actuator
Chapter 6. Computation of the System

Figure 59 - Displacement for actuator Y3

Figure 60 - Displacement for actuator Y4

Chapter 6. Computation of the System

Displacement actuator X1 – X2

Input: displacement X direction, displacement Y direction.

∆X actuator
=actuator displ . component ∈X direction=displ .record X direction

∆Y actuator
=actuator displ . component ∈Y direction=displ .record Y direc tion

∆Y + L0 ∗sin ( ϑ0
ϑactuator =atan
[ L0
actuator

actuator
∗cos ( ϑ0
actuator

actuator
actuator

)+ ∆ X
)
actuator
]
−ϑ 0 actuator

L0 ∗cos ( ϑ0 )+ ∆ X
∆ actuator =
[ actuator

cos ( ϑ actuator +ϑ0 )

actuator

actuator
actuator

] −L0 actuator

Input: displacement X direction, displacement Y direction, yaw rotation.

ϑtriangle = ϑcentroid − yaw rotation → for actuator X 1

{
ϑ centroid+ yaw rotation→ for actuator X 2

∆X actuator
=− [ ∆ R∗sin ( ϑtriangle )−∆ X ] + displacement X direction

−[ ∆ R∗cos ( ϑtriangle )−∆Y ]+ displacement Y direction → for actuator X 1

∆Y actuator
=
{[ ∆ R∗cos ( ϑtriangle )−∆Y ]+ displacement Y direction → for actuator X 2

∆Y + L0 ∗sin ( ϑ0
ϑactuator =atan
[ L0
actuator

∗cos ( ϑ0
actuator
actuator

actuator
)+ ∆ X
actuator
)
actuator
]
−ϑ 0 actuator

L0 ∗cos ( ϑ0 )+ ∆ X
∆ actuator =
[ actuator

cos ( ϑ actuator +ϑ0 )

actuator

actuator
actuator

] −L0 actuator
Chapter 6. Computation of the System
Chapter 6. Computation of the System

Figure 61 - Displacement for actuator X1

Figure 62 - Displacement for actuator X2

Chapter 6. Computation of the System

Displacement actuator X3 – X4

Input: displacement X direction, displacement Y direction.

∆X actuator
=actuator displ . component ∈X direction=displ .record X direction

∆Y actuator
=actuator displ . component ∈Y direction=displ .record Y direction

∆Y ∗sin ( ϑ0
ϑactuator =−atan
[ L0
actuator

actuator
−L0
∗cos ( ϑ0
actuator

actuator
actuator

) −∆ X
)
actuator
] −ϑ0 actuator

L0 ∗cos ( ϑ0
∆ actuator =
[ actuator
) −∆ X
actuator

cos ( ϑactuator +ϑ0 ) actuator

actuator

] −L0 actuator

Input: displacement X direction, displacement Y direction, yaw rotation.

ϑtriangle = ϑcentroid + yaw rotation → for actuator X 3

{
ϑcentroid − yaw rotation → for actuator X 4

∆X actuator
=[ ∆ R∗sin ( ϑ triangle ) −∆ X ] +displacement X direction

−[ ∆ R∗cos ( ϑtriangle )−∆Y ]+ displacement Y direction → for actuator X 3

∆Y actuator
=
{[ ∆ R∗cos ( ϑ triangle ) −∆Y ] +displacement Y direction → for actuator X 4

ϑactuator =−atan ¿ ¿

L0 ∗cos ( ϑ0
∆ actuator =
[ actuator
) −∆ X
actuator

cos ( ϑactuator +ϑ0 ) actuator

actuator

] −L0 actuator
Chapter 6. Computation of the System

Figure 63 - Displacement for actuator X3

Figure 64 - Displacement for actuator X4

Chapter 6. Computation of the System

Figure 65 - Rotation of the actuators in Y direction with no rotational input for the system (red line)

Figure 66 - Rotation of the actuators in X direction with no rotational input for the system (red line)
Chapter 6. Computation of the System

6.1.2 Force
The calculation of the force at each instant of the input records requires the evaluation of the
moving mass, taking into account also the mass of the actuators in the horizontal and vertical
direction.

1
M X =M shake table + M specimen + ∗( M Y 1+ M Y 2+ M Y 3 + M Y 4 + M Z 1+ M Z 2)
3

1
M Y =M shake table + M specimen + ∗( M X 1+ M X 2+ M X 3+ M X 4 + M Z 1 + M Z 2 )
3

Then, the force is simply calculated as:

F X (t)=M X∗Acc X (t )

F Y (t )=M Y ∗AccY (t)

For the mass of the actuators, 1 ton has been considered. Then, the shake table reaches with
the last plate configuration 52.71 tons and the maximum allowable specimen mass is equal to
120 tons.

Figure 67 - Force for the component X and Y of the motion

Chapter 6. Computation of the System

6.1.3 Hydraulic system (flow)

The flow rate for each of the horizontal actuators is evaluated considering the absolute value
of the actuator velocity input and the piston area:

Flowi ( t )=|Velocity( t)|∗PistonArea

By summing up all the contributions of the flows corresponding to the 8 actuators is possible
to calculate the total flow rate:

TotalFlowRate ( t )=∑ Flow i (t)

The obtained total flow rate is then compared with the continuous flow rate given by the
pumping system and is equal to 360 l/min. This is actually the flow rate addressed to the new
facility through the pumping system which is shared with the existent shake table and is the
amount of oil given by 2 pumps (180 l/min capacity each).

Finally the calculation of the total flow volume is done by integrating the total flow rate. The
numerical integration uses the trapezoidal rule.
Chapter 6. Computation of the System

Figure 68 - Flow rate of the actuators in Y direction

Chapter 6. Computation of the System

Figure 69 - Flow rate of the actuators in X direction

Figure 70 - Total flow rate comparison

Chapter 6. Computation of the System

6.1.4 Accumulators demand

The evaluation of the accumulators demand is done by considering the constant flow rate
given by the pumping system. The value considered is 360 l/min and this value corresponds to
the partial pumping system capacity of TREES Lab that is shared together with the hydraulic
system of the new facility. In fact, the hydraulic system for the new shake table will use the
existing pumping system and a new high capacity accumulation system has to be designed.

The calculation of the demand has been performed by “cutting” the continuous flow rate from
the entire flow rate demand. The integration of the flow rate above the value of 360 l/min
gives the demand for the accumulation system in terms of flow volume.

Figure 71 - Shifted flow rate comparison

For each of the selected records, the total flow volume, the accumulator demand and the
contribution to the total amount of oil given by the pumping system are compared like in the
following picture.
Chapter 6. Computation of the System

Figure 72 - Total flow volume comparison

The entire set of earthquakes records have been selected using the PEER NGA Database
(http://peer.berkeley.edu/nga/earthquakes.html). In order to design the shake table system, and
especially the hydraulic part of it, the flow calculations have been made considering
earthquakes with magnitude higher than 6, distances up to 20 km in order to take into
account for the near fault effects and PGA values between 0.2 g and 2 g.

The data used in the MATLAB sheet for the specific calculations are:

- input acceleration X, acceleration Y;

- baseline correction applied;
- acceleration scale factor 1;
- L0actuator 3.0836 m;
- ϑ0actuator 0°;
- ϑcentroid 57.41°;
- distance between the two actuators (same side) 4.8 m;
- piston area 75.6 in2;
- actuator mass 1 ton;
- specimen mass 120 tons.
- continuous flow rate 360 l/min.
Chapter 6. Computation of the System

Record Earthquake Station MW Rclosest [km] PGA [g] PGV [cm/s] PGD [cm] Δt [sec]
NGA007 San Fernando 1971-02-09 14:00 CDMG 279 Pacoima Dam (upper left 6.61 1.81 1.16 75.55 18.06 0.0100
NGA045 Morgan Hill 1984-04-24 21:15 CDMG 57217 Coyote Lake Dam (SW 6.19 0.53 0.97 68.35 10.21 0.0050
NGA072 Superstition Hills-02 1987-11-24 USGS 286 Superstition Mtn Camera 6.54 5.61 0.79 36.89 5.84 0.0100
NGA075 Loma Prieta 1989-10-18 00:05 CDMG 57007 Corralito 6.93 3.85 0.50 41.92 10.12 0.0050
NGA076 Loma Prieta 1989-10-18 00:05 CDMG 47381 Gilroy Array #3 6.93 12.82 0.46 43.11 11.83 0.0050
NGA082 Cape Mendocino, 1992-04-25 18:06 CDMG 89324 Rio Dell Overpass - FF 7.01 14.33 0.42 47.95 16.96 0.0200
NGA100 Northridge-01 1994-01-17 12:31 USGS/VA 637 LA - Sepulveda VA 6.69 8.44 0.80 74.13 16.32 0.0050
NGA101 Northridge-01 1994-01-17 12:31 DWP 0 LA Dam 6.69 5.92 0.45 53.70 17.05 0.0050
NGA108 Northridge-01 1994-01-17 12:31 CDMG 24514 Sylmar - Olive View 6.69 5.30 0.70 95.38 21.94 0.0200
NGA110 Kobe, Japan 1995-01-16 20:46 (6.9) JMA 99999 KJMA 6.90 0.96 0.71 77.83 18.87 0.0200
NGA111 Kobe, Japan 1995-01-16 20:46 (6.9) CUE 99999 Takarazuka 6.90 0.27 0.71 75.88 23.13 0.0100
NGA116 Kocaeli, Turkey 1999-08-17 (7.51) ERD 99999 Izmit 7.51 7.21 0.20 27.02 14.61 0.0050
NGA118 Chi-Chi, Taiwan 1999-09-20 (7.62) CWB 99999 CHY006 7.62 9.77 0.36 52.07 20.95 0.0040
NGA119 Chi-Chi, Taiwan 1999-09-20 (7.62) CWB 99999 CHY028 7.62 3.14 0.79 71.98 18.19 0.0050
NGA122 Chi-Chi, Taiwan 1999-09-20 (7.62) CWB 99999 CHY074 7.62 10.80 0.21 26.89 14.88 0.0050
NGA149 Chi-Chi, Taiwan 1999-09-20 (7.62) CWB 99999 TCU055 7.62 6.36 0.22 36.87 22.02 0.0050
NGA152 Chi-Chi, Taiwan 1999-09-20 (7.62) CWB 99999 TCU088 7.62 18.16 0.52 19.06 13.14 0.0050
NGA160 Duzce, Turkey 1999-11-12 (7.14) ERD 99999 Bolu 7.14 12.04 0.77 59.68 17.69 0.0100
NGA161 Duzce, Turkey 1999-11-12 (7.14) LAMONT 1062 Lamont 1062 7.14 9.15 0.21 13.70 9.43 0.0100
NGA161 Duzce, Turkey 1999-11-12 (7.14) LAMONT 375 Lamont 375 7.14 3.93 0.74 28.24 6.09 0.0100
Table 15 - Set of earthquake records for the flow requirements calculations

The comparison of the results in terms of total flow rate and total flow volume for each of the record will lead to a better understanding of
what it could be the possible requirements for the capacity of the accumulation system. The following results show that a capacity of 1000
liters allows performing the 85% of the record set. This value is considered as a reasonable value for the capacity of the accumulators.
Chapter 6. Computation of the System

Figure 73 - Acceleration records for the X direction

Figure 74 - Acceleration records for the Y direction

Chapter 6. Computation of the System

Figure 75 - Spectra of the acceleration records in the X direction

Figure 76 - Spectra of the acceleration records in the Y direction

Chapter 6. Computation of the System

Figure 77 - Total flow rate demand for the set of acceleration records
Chapter 6. Computation of the System

Figure 78 - Total flow volume demand for the set of acceleration records

Figure 79 - Flow volume demand for the accumulation system and accumulator capacity
Chapter 6. Computation of the System

6.2 Specifications for the shaking table design

The construction technology is an important choice related to the way in which the system can
be fabricated in terms of materials, their assemblage, etc. The manufacturing technology
should guarantee the best performance behavior of the table and of its components. The top
plate has to be provided with a regular grid of holes in plan to secure the test specimen. It
should also be able to resist deformation during testing. Precision tolerances have to be
included within acceptable ranges to ensure smooth and plane surfaces.

The choice of the construction material is limited by the specifications of the facility. If the
table must be light, aluminium gives the possibility to create limited weight platforms. If the
customer wants a heavy platform, steel is the most appropriate material with the possibility of
adding concrete blocks within the table to reach the design weight. A very light platform has
the advantage that it requires a smaller actuator to move it, which thereby reduces the initial
cost of the facility. However, a very light platform is much more likely to be affected by
significant table-specimen interaction which requires better hardware and software systems to
effectively control. Therefore the larger the platform mass the better it is from the point of
view of the specimen, but this requires larger actuators, increases running costs and can limit
the maximum accelerations achievable. A solution of this conflict is the use of a light
platform with high actuators capacity. For small light specimens the platform is used as it is,
but when a larger specimen that may interact significantly with the table is to be tested, then
additional static mass, up to the capacity of the table, may be added to the platform. This
additional mass then helps to reduce the table-specimen interaction by increasing the platform
mass that has to be excited by the specimen. The other effect of having a larger, more massive
platform is simply the enhanced capacity of the shaking table to deal with larger scale models.
The larger the model, the simpler the scaling issue becomes. Therefore, within the limits
imposed by cost of construction of the shaking table, the running costs and the cost of
manufacture of models, the largest mass of the platform and capacity of the table should be
chosen.

The stiffness of the table and the bearings must be very high to reach frequency responses
beyond the range of fundamental frequencies of the test structures. The design of a shaking
table requires accurate considerations of the response frequencies of all the components of the
facility since the resonance of the response can cause unforeseen amplifications of motions.
The platform should be sufficiently rigid in order not to respond itself dynamically so that it
transmits the input motion to the structure with as little modification as possible. In addition,
there is the variation in motion across the table surface itself.
Chapter 6. Computation of the System

6.2.1 Dynamic response characteristics

A full finite element model is created in order to evaluate the eigenfrequencies of the shake
table. Some trials were done manually changing the thickness of the plates to get a high first
eigenfrequency.

The initial model was created by considering the layout of the first draft in which the 4
horizontal actuators were connected to the shaking table along the web being in this way
hidden by the shaking table itself. The following figures show the plate alignments and the
position (colored areas) of the horizontal actuators, the vertical hold-downs and the sliding
bearings.

Figure 80 - Vertical alignments (longitudinal and transversal) x-y plane (measures in cm) - First draft
Chapter 6. Computation of the System

Alignmen
Plane Elements Distance [cm] Area element [cm2] Thickness [cm]
t
140 1h 490000 3
Horizontal 100 2h 490000 2
x-y
layers 60 3h 490000 3
0 4h 113400 3
0 1t 71240 1
30 2t 71240 1
110 3t 71240 1
190 4t 71240 1
270 5t 71240 1.5
310 6t 71240 1.5
Transversal
x-z 350 7t 71240 1.5
layers
390 8t 71240 1.5
430 9t 71240 1.5
510 10t 71240 1
590 11t 71240 1
670 12t 71240 1
700 13t 71240 1
0 1l 51800 1.5
30 2l 51800 1.5
110 3l 51800 1.5
190 4l 51800 1.5
270 5l 133700 1.5
Longitudina
y-z 350 6l 133700 1.5
l layers
430 7l 133700 1.5
510 8l 51800 1.5
590 9l 51800 1.5
670 10l 51800 1.5
700 11l 51800 1.5
Table 16 - Thickness of the shake table plates (first draft)
Chapter 6. Computation of the System

Figure 81 - Horizontal alignments y-z plane (measures in cm) - First draft

Figure 82 - Horizontal alignments x-z plane (measures in cm) - First draft

In Table 16 the specifications for all the alignments are showed. With this configuration the
shake table reaches 52.71 ton. This value, considering the table dimension of 7 by 7 m, can be
considered as a good preliminary agreement between table stiffness and weight.

The carried out modal analysis of the first draft finite element model exhibited the following
results.

Figure 83 - View of the shake table model and restraints (first draft)
Chapter 6. Computation of the System

Mode Period [sec] Frequency [Cyc/sec] Circ. Frequency [rad/sec]

1 0.009574 104.45 656.29
2 0.009135 109.46 687.79
3 0.006013 166.3 1044.9
4 0.005783 172.91 1086.4
5 0.005543 180.41 1133.6
6 0.005543 180.41 1133.6
7 0.00553 180.85 1136.3
8 0.005529 180.85 1136.3
9 0.005458 183.23 1151.3
10 0.005458 183.23 1151.3
11 0.005444 183.67 1154.1
12 0.005444 183.68 1154.1
13 0.00544 183.84 1155.1
14 0.005439 183.86 1155.2
15 0.00543 184.17 1157.2
16 0.00543 184.18 1157.2
17 0.005415 184.66 1160.2
18 0.005415 184.66 1160.3
19 0.005386 185.66 1166.5
20 0.005385 185.69 1166.7
Table 17 - Result of the modal analysis (first draft)

Figure 84 - First mode shape 104 Hz (first draft)

Chapter 6. Computation of the System
Chapter 6. Computation of the System

Figure 85 - Second mode shape 109 Hz (first draft)

Figure 86 - Third mode shape 166 Hz (first draft)

Chapter 6. Computation of the System

Figure 87 - Fourth mode shape 72 Hz (first draft)

Mode SumX SumY SumZ SumRX SumRY SumRZ

1 2.636E-11 1.263E-11 0.58 0.41 0.41 1.115E-12
2 2.7E-11 0.02909 0.58 0.59 0.41 0.01103
3 2.202E-09 0.02909 0.62 0.62 0.44 0.01103
4 5.746E-09 0.85 0.62 0.64 0.44 0.32
5 1.68E-08 0.85 0.62 0.64 0.44 0.32
6 2.294E-08 0.85 0.62 0.64 0.44 0.32
7 2.378E-08 0.85 0.62 0.64 0.44 0.32
8 2.594E-08 0.85 0.62 0.64 0.44 0.32
9 2.706E-08 0.85 0.62 0.64 0.44 0.32
10 3.155E-08 0.85 0.62 0.64 0.44 0.32
11 3.187E-08 0.85 0.62 0.64 0.44 0.32
12 3.188E-08 0.85 0.62 0.64 0.44 0.32
13 0.003785 0.85 0.62 0.64 0.44 0.33
14 0.003785 0.85 0.62 0.64 0.44 0.33
15 0.003817 0.85 0.62 0.64 0.44 0.33
16 0.003817 0.85 0.62 0.64 0.44 0.33
17 0.003817 0.85 0.62 0.64 0.44 0.33
18 0.003817 0.85 0.62 0.64 0.44 0.33
19 0.0107 0.85 0.62 0.64 0.44 0.33
20 0.0107 0.85 0.62 0.64 0.44 0.33
Table 18 - Participating masses (first draft)
Chapter 6. Computation of the System

The results of the modal analysis satisfy the requirements for the stiffness specification at
least at this preliminary design level. Although the stiffness could be considered sufficiently
high, the layout with the actuators inclined and attached to the web of the table doesn’t allow
a high performance in terms of frequency. Another solution has been studied and compared
with the previous one.

In the second draft, 8 horizontal actuators are moving the shake table, 2 for each table side.
The table is constraint to the horizontal actuators, the vertical hold-downs and the bearing
pads, considering a hinge connection. There is no need to perform several modal analysis
corresponding to all the possible restraint configurations of the table cause the restraint
configuration is unique due to the fact that the bearing pads are fixed to the bottom of the
shaking table and not to the sliding surface of the foundation mass. The following figures
show the plate alignments and the position (colored areas) of the horizontal actuators, the
vertical hold-downs and the sliding bearings.

Figure 88 - Vertical alignments (longitudinal and transversal) x-y plane (measures in cm) - Second draft
Chapter 6. Computation of the System

Figure 89 - Horizontal alignments y-z plane (measures in cm) - Second draft

Figure 90 - Horizontal alignments x-z plane (measures in cm) - Second draft

The carried out modal analysis of the second draft finite element model exhibited the
following results. Only the first four modes can be consider as global modes involving a
considerable table mass.

Figure 91 - View of the shake table model and restraints (second draft)
Chapter 6. Computation of the System

Figure 92 - View of the inner layer scheme (second draft)

Mode Period [sec] Frequency [Cyc/sec] Circ. Frequency [rad/sec]

1 0.007334 136.35 856.71
2 0.00685 145.98 917.21
3 0.006324 158.12 993.49
4 0.006272 159.43 1001.7
5 0.005549 180.22 1132.4
6 0.005549 180.22 1132.4
7 0.005528 180.91 1136.7
8 0.005527 180.91 1136.7
9 0.005461 183.12 1150.6
10 0.005461 183.12 1150.6
11 0.005446 183.62 1153.7
12 0.005446 183.63 1153.8
13 0.005445 183.66 1154
14 0.005445 183.66 1154
15 0.005428 184.23 1157.6
16 0.005428 184.23 1157.6
17 0.005426 184.31 1158
18 0.005426 184.31 1158.1
19 0.005391 185.5 1165.5
20 0.00539 185.53 1165.7
Table 19 - Result of the modal analysis (second draft)
Chapter 6. Computation of the System

Figure 93 - First mode shape 136 Hz (second draft)

Figure 94 - Second mode shape 145 Hz (second draft)

Chapter 6. Computation of the System

Figure 95 - Third mode shape 158 Hz (second draft)

Figure 96 - Fourth mode shape 159 Hz (second draft)

Chapter 6. Computation of the System

The higher modes involve a low participating mass cause they are due to the local modes of
the inner plates. In the following table the participating mass for displacement X,
displacement Y, displacement Z (vertical), rotation about X axis (pitch rotation), rotation
about Y axis (roll rotation) and rotation about Z (yaw rotation) are shown.

Mode SumX SumY SumZ SumRX SumRY SumRZ

1 7.467E-12 5.927E-12 0.57 0.41 0.4 2.055E-13
2 1.267E-11 0.02368 0.57 0.55 0.4 0.008937
3 0.03773 0.02368 0.57 0.55 0.55 0.02317
4 0.03773 0.02368 0.68 0.63 0.63 0.02317
5 0.03773 0.02369 0.68 0.63 0.63 0.02317
6 0.03773 0.02369 0.68 0.63 0.63 0.02318
7 0.03773 0.02369 0.68 0.63 0.63 0.02318
8 0.03773 0.02369 0.68 0.63 0.63 0.02318
9 0.03773 0.02369 0.68 0.63 0.63 0.02318
10 0.03773 0.02369 0.68 0.63 0.63 0.02318
11 0.03951 0.02369 0.68 0.63 0.63 0.02386
12 0.03951 0.02369 0.68 0.63 0.63 0.02386
13 0.03951 0.02369 0.68 0.63 0.63 0.02388
14 0.03951 0.02369 0.68 0.63 0.63 0.02388
15 0.03951 0.02369 0.68 0.63 0.63 0.02391
16 0.03951 0.02369 0.68 0.63 0.63 0.02391
17 0.03953 0.02369 0.68 0.63 0.63 0.02392
18 0.03953 0.02369 0.68 0.63 0.63 0.02392
19 0.04402 0.02369 0.68 0.63 0.63 0.02561
20 0.04402 0.02369 0.68 0.63 0.63 0.02561
Table 20 - Participating masses (second configuration)

The highest values of participating mass are for the vertical displacement (68% of the total
table mass) and the rotations about the two horizontal axes (63 % of the total table mass for
both rotations). As previously written, it’s clear that only the first four modes involve a
considerable table mass while all the other modes are local modes involving the inner plates.
Chapter 6. Computation of the System

6.2.2 Overturning moment

The position of the actuator connection influences the overturning moment effect. If the
actuator connection is below the centre of gravity of the table, inertial forces from the
specimen and the table masses act together creating overturning moment at the bearings.
Contrarily, if the centre of gravity of the table is below the actuator connection joint, the
inertial force of the table gives an opposing moment thereby reducing the overall overturning
moment.

The position of the centre of gravity for the case study was calculated. Considering that the
main references are on the bottom horizontal plate, the coordinates of the centre of gravity are
(the horizontal plane corresponds to the x-y plane):

xG = 3500 mm yG = 3500 mm zG = 904 mm

The value in the vertical direction is meant to be at around 90 cm above the bottom horizontal
plate. In the drawing below it’s possible to see the position of the centre of gravity, the
alignment of the plates and the position in which the horizontal actuators, the vertical hold-
down and the bearing pads are attached to the table (rectangular areas). The eccentricity of the
horizontal actuators in the vertical direction is about 10 cm.
Chapter 6. Computation of the System

Figure 97 - Position of the centre of gravity and position of the layers (measures in cm)
Chapter 6. Computation of the System

6.3 Specifications for the system restraining the platform motion

The hydrostatic bearings represent physical restraints required to prevent motion in the
unwanted degrees of freedom where active control is not provided on all axes. However there
is a risk that such a restraint system affects the motion in the actively controlled directions.
The bearings should give the directionality of the motion and compensate the effects of the
overturning moment in the longitudinal and transversal directions. The main idea is to allow
sliding of the shaking table in the horizontal directions. Their dimensions, number and
locations are chosen in order to allow the desired motion and to avoid the undesired ones.

Load paths in the bearings and consequently their deformations must be evaluated. As
introduced before, the bearings must be manufactured with high precision. The tolerance error
has to be negligible and they cannot deform to an extent greater than the viscous fluid
thickness. The viscous fluid is located on the upper and lower surfaces of each of the lateral
bearings and on the external sides of the bottom bearings. The deformations of table and
bearings should allow the stroke of the platform without causing mechanical problems
between the bearings and the fixed guide system. The thickness of the viscous fluid film, h,
can be in the range of 0.1 mm to 0.02 mm. The thicker is the viscous fluid film the stiffer is
the provided restraint.

6.4 Specifications for the reaction mass design

The table’s resultant motion is transmitted to the reaction mass, to the isolators and dampers,
then to the foundation and finally, to the underlying soil. The reaction mass is constructed in
reinforced concrete and should have a typical mass ratio 30 to 50 times the specimen-table
mass in order to reduce the table’s resultant motion to the very minimum. The concrete block
could be isolated from the rest of the laboratory and the underlying structures by inflating air
suspension springs under the block. The reaction block has dimensions generally imposed by
the choice of the structural layout and the soil characteristics. Large forces are transmitted to
the reaction mass. As a first possibility, the need for a rigid structure with high performance
and constructional accuracy could be fulfilled by precast concrete blocks assembled by post-
tensioning tendons. A totally different design could consist of casting in-situ reinforced
concrete. With this solution, the designer could increase the mass ratio more easily. The
disadvantages of this solution could be the construction costs, the implication of new costs for
the excavation of a large volume of soil, the possibility of reaching the water table level and
the interaction with the foundation system of the laboratory, which could be close to the
reaction mass perimeter.

The resultant motion of the table is frequency dependent and it is “filtered” by the frequency
response of the reaction mass, the isolators and dampers and the foundation. Accurate
analyses must be performed in order to solve the dynamic soil-structure interaction problem
and predict the ground motion induced by the excitation of the table.
Chapter 6. Computation of the System

6.5 Specifications for the isolation and damping system design

The isolation system is aimed at moving the response period of the reaction mass far from the
specimen’s fundamental response periods and in a range where the actuator performances are
small. The isolation of the dynamic facility is a design choice when the site conditions do not
allow the reaction mass to rest directly on the supporting soil. The use of the isolation system
has the advantage to reduce the motion transmitted to the soil and to mitigate the vibrational
impact of the tests. High values of peak acceleration transmitted to the foundation/soil can be
potentially hazardous for structural damage in nearby buildings and in the laboratory itself.
Furthermore the foundation soil may be subjected to high strain levels that can invalidate the
results obtained from loading tests due to the loss of coherence caused by spurious vibrations.
The isolation system should be characterized by a fundamental frequency and damping ratio
such that the amplitude of the transmitted motion within the operational frequency range is
minimized. Due to the necessity of reducing the transmitted motion to the maximum, the
isolators must be characterized by small horizontal stiffness and high damping ratio.

Figure 98 - Components of a typical shake table Figure 99 - Components of a typical shake table
(plan view) [Clark, 1992] (section) [Clark, 1992]

The dynamic performance of the shaking table system can be significantly affected by
potential regions of flexibility:

- the flexibility of the reaction mass on the suspension system/shock absorbers system;
- the reaction mass internal flexibility;
- local flexibility of support brackets on reaction mass;
- flexibility and any backlash in the actuator bearings
- axial and lateral bending stiffness of the actuators;
- hydraulic oil column bulk modulus stiffness;
- axial, torsional and lateral bending stiffness of any torsion tubes or other restraining
system;
- flexibility of the platform.
Chapter 7. Conclusions

7 CONCLUSIONS
The aim of this work is to present the preliminary studies and analyses that have been
developed in order to define the specifications for the new laboratory of TREES Lab 2 at
EUCENTRE. The reasons below the needs of having such a facility are due to the increase, in
the last decades, of more demanding experimental research dealing with big scale models and
strong acceleration records.

It has to be pointed out that defining the characteristics of an advanced research laboratory in
the field of earthquake engineering corresponds in satisfying the requirements of modularity,
flexibility and complementarity considering the existing research centres. The number of new
high performance shake table facilities is increasing all around the world and this would soon
restrict the European continent to a secondary role in the field of earthquake engineering.
With this new laboratory, Europe will be equipped with a new world class facility for
dynamic and pseudo-dynamic testing

A detailed layout of the facility is proposed. Several and strong acceleration records are
considered for the design of the high performance accumulation system and a value of
capacity has been set according to the performances required for TRESS Lab 2. This proposal
will need further investigations to define more precisely the details of the tables’ plates, the
hydraulic system, the foundation mass and other complementary aspects necessary for
successful experiments, such as, measurement techniques, technology and control issues.
References

8 REFERENCES
Calvi, G.M.,Pavese, A., Ceresa, P., Dacarro, F., Lai, C.G., Beltrami, C. [2005] Design of a large-scale
dynamic and pseudo-dynamic testing facility, IUSS Press, Pavia, Italy.
Ceresa, P. [2004] “Design of a dynamic and pseudo-dynamic testing facility” MsC thesis, European
School for Advanced Studies in Reduction of Seismic Risk (ROSE School), University of Pavia,
Italy.
Chopra, A.K., [2001] Dynamics of structures, Second Edition, Prentice Hall, Upper Saddle River,
United States of America.
Van Den Einde, L., Restrepo, J., Conte, J., Luco, E., Seible, F., Filiatrault, A., Clark, A., Johnson, A.,
Gram, M., Kusner, D. and Thoen, B. [2004] “Development of the George E. Brown Jr. Network for
Earthquake Engineering Simulation (NEES) Large High Performance Outdoor Shaking Table at
the University of California, San Diego” Proceedings of 13th World Conference on Earthquake
Engineering, Vancouver, British Columbia, Canada, paper 3281.
Ohtani, K., Ogawa, N., Katayama, T., Shibata, H. [2003] “World’s Largest Shaking Table Takes
Shapes in Japan” Transactions of the 17th International Conference on Structural Mechanics in
Reactor Technology, Prague, Czech Republic.
Project EFAST [2010], “Design Study of a European Facility for Advanced Seismic Testing” 7th
Framework Programme Capacity-Research Infrastructures, Deliverable Report.
Marazzi, F., Politopoulos, I., Pavese, A., [2010] “An overview of seismic testing needs in Europe:
towards a new advanced experimental facility”, Bull Earthquake Eng (2011) 9:623–640.
Appendix A

APPENDIX A

Figure 100 - Plan view of the two laboratories at EUCENTRE

A1
Appendix A

Figure 101 - Section A_A of TREES Lab 2 structure

Figure 102 - Section B_B of TREES Lab 2 structure

Appendix A

Figure 103 - Plan view of the laboratories and the buildings of University of Pavia

A3
MSc Dissertation 2011Study and development of a high performance shake table system - TREES Lab 2Alessandro Scodeggio
Appendix A