Seismic Engineering Research Infrastructures for European Synergies. JRA1: Small Lab Experience. Iasi 13th July 2009

Transcription

1 Seismic Engineering Research Infrastructures for European Synergies Structural Dynamics Research Group Department of Engineering Science University of Oxford JRA1: Small Lab Experience Iasi 13th July 2009

2 Summary Involvement with JRA1 & 2 JRA1 novel actuators TT1 testing (real time testing) Testing & calibration of load cells & actuators Data formats Laboratory overview Structural Dynamics Laboratory Testing installation Control installation Testing experience 2

3 Novel devices & sensors JRA1 Development of novel actuators electro-magnetic linear motors JRA2 TT1 testing (real time testing with UNITN) Test method to assess novel sensors in more realistic environment Use both hydraulic and electromagnetic actuators Data formats (JRA2 & NA1) 3

4 Index Laboratory overview Structural Dynamics Laboratory Testing installation Control installation Videoconference & Telepresence installation Current testing 4

5 Laboratory overview Structural Dynamics Laboratory Located in Jenkin Building at University of Oxford. It has a reinforced concrete strong floor isolated from the laboratory building. The laboratory is primarily set up for hybrid testing, real-time hybrid testing on linear, nonlinear and rate dependent physical components The laboratory overview will be divided in three main parts: Main testing installation Control installation Grandstand rig 5

6 Index Laboratory overview Structural Dynamics Laboratory Testing installation Control installation Testing Experience 6

7 Main Structural Dynamics Laboratory 7

8 Laboratory overview (cont) Testing installation The main hydraulic installation consists of: 3 pumps An oil reservoir 5 accumulators 2 Instron substations 6 Instron actuators 8

9 Laboratory overview (cont) Testing installation The hydraulic supply 3 hydraulic pumps 270 l/min 210 bar 5 accumulators additional 180 l 6 Instron dynamic servo-hydraulic actuators Moog servovalves 2 off 250 kn ±125mm 2 off 100 kn ±125mm 2 off 10 kn ± 50mm 9

10 Grandstand test rig (separate lab) UK EPSRC funded study 10

11 Electro-magnetic linear motors Testing installation 4 off Oswald electrical actuators 10 kn peak load 120 mm stroke 4 off Danaher inverters Oswald Electrical actuators 11

12 Index Laboratory overview Structural Dynamics Laboratory Testing installation Control installation Testing experience 12

13 Laboratory overview (cont) Control installation All the equipment in the main lab is controlled from the service deck: Testing Telepresence Videoconferencing 13

14 Laboratory overview (cont) Control installation The main control devices are: dspace DS1104 card, to control experiments and capture data dspace Host PC, a quad core machine with dual monitor to serve the dspace board Instron 8800 controller UK-light switch (CISCO) high-speed network connection to Bristol and Cambridge Telepresence server, running Red Hat Enterprise Linux 14

15 Laboratory overview (cont) Control installation Instron Controller is connected to two computers: dspace host PC Instron Controller PC 15

16 Index Laboratory overview Structural Dynamics Laboratory Control installation Testing installation Testing experience 16

17 Testing experience Real-time (fast) hybrid testing 3 dof inertial rig Hysteretic & viscous dampers 2 dof stiffness rig 17

18 Hybrid testing: displacement control loop Apply measured forces to numerical substructure External loads (eg. earthquake) Physical substructure Numerical substructure Measure forces and actual displacements of test specimen Control lag/ delay Communication/ calculation delay Calculate displacements at interface between physical and numerical substructures Command actuators to apply interface displacements to physical substructure 18

19 Real-time hybrid testing Accurate, stable, fast tests depend on smallest possible lag/delay in control loop minimal filtering Responsive actuators Low-noise sensors 19

20 Typical response: hysteretic sample Force (kn) Force demand Measured force Error Time (s) Good response (in force control): Demand well matched with ~10ms delay 20

21 Non-linear viscous sample Less good when sudden high displacements are needed High flow demands cause lag in response Force demand Measured force Error Force (kn) Time (s) 21

22 Electro-magnetic actuators Grandstand simulation rig 15 seat seating deck Oswald electro-magnetic linear motors (xpc Target control) 22

23 Linear motor Motor is powered by an inverter Feedback signal from encoder, which encoder uses to commutate the current to the windings 415V 3-phase supply is rectified to 600V DC and switched to motor windings at 16kHz High frequency alternating e/m field is ejected along the axis of the motor 23

24 Problems Low force capacity 20kN maximum Prefer zero mean load take out mean load by other means 24

25 Load cell on back left actuator Measured force (offset) Load demand Load (N) Time (s) 25

26 Electro-magnetic actuators Motors Good but limited capacity Require large currents Inverters Highest capacity motors limited range available Aimed at industrial use poor noise performance on monitoring signals Sensors High e/m fields cause interference better instrumentation needed 26

27 Possible area for improving sensors Linear motors cause intense e/m fields close to the device Standard strain gauge load cells attached to actuators suffer significant interference High speed testing requires clean accurate feedback signals Need research on sensors to use e/m (& servo-hydraulic) actuators effectively 27

28 Electro-mechanical actuators Little experience at Oxford (soils lab uses them) Ball-screw drive Higher loads up to 44kN Large stroke Control?? (Source: Parker catalogue) Trento will be investigating these in JRA2 28