Case Study : Yokohama-Bay Bridge

Transcription

1 Case Study : Yokohama-Bay Bridge D3-X,D3-Y,D3-Z D6-YL,D6-ZL D8-YL,D8-ZL D1-X,D1-Y,D1-Z D7-X,D7-Y,D7-Z D9-X,D9-Y,D9-Z D5-X,D5-Y,D5-Z D2-Y,D2-Z D4-Y,D4-Z D6-YR,D6-ZR D8-YR,D8-ZR 200 m 460 m 200 m T4-X, T4-Y T3-XL, T3-YL T3-XR, T3-YR T8-X, T8-Y T7-XL, T7-YL T7-XR, T7-YR P1-X, P1-Y Input Channel : 1,2,3 T2-X, T2-Y T1-XL, T1-YL Input Channel : 4,5,6 T1-XR, T1-YR T6-X, T6-Y T5-XL, T5-YL Input Channel : 7,8,9 T5-XR, T5-YR P2-X, P2-Y Input Channel : 10,11,12 Truss-box girder, double-deck cable-stayed bridge Completed in 1988 Span length: m, Tower : 172 m x 29 m Permanently instrumented with 85 channels of accelerometers

2 Observed Longitudinal Mode from Sys. Id Three observed typical first longitudinal modes : Relative Modal Displacement between Pier and Girder (ψ): ϕ = φ φ girder pier cap φ girder φ pier cap Fix Connection Left_end Right_end Free Connection Left_end Right_end (b) Typical (c) Typical Mixed Hinged-Fixed Fixed-Fixed (a) Typical Mode Hinged-Hinged Mode Mode (Earthquake (Earthquake Frame-1) Frame-2) (Earthquake Frame-1)

3 Advantages of WSN for SHM applications Wireless sensors potentially provides: Inexpensive & Dense measurement; Easy installation Autonomous measurement & judgment Key to comprehend structural performance 3

4 WSN system development Applications data aggregation Middleware networking sensing Synchronized sensing (Multihop) Ack based efficient data transfer RF CPU Memory OS Power Hardware Sensor/ actuator TinyOS Imote2 (MEMSIC)+ customized sensor boards Imote2: CPU 400MHz RAM 32MB RF: MEMS Accelerometer + ADC

5 Seoul Daejeon (KAIST) Full scale bridge vibration measurements: Jindo Bridge, Korea (September 2008 present) Jindo Jindo Bridges Jeju Island 2 nd nd Jindo Bridge Haenam (Inland) Jindo Island The 2 nd Jindo Bridge Type Spans Girder Design velocity Designed by Constructed by Owner Special feature Cable-stayed bridge = 484m Steel box (12.55m width) 70 km/hr Yooshin cooperation (2000, Korea) Hyundai construction (2006, Korea) Iksan Regional Construction and Management Administration Twin bridge 5

6 Acceleration (mg) Sensor validation preliminary test Wired and wireless sensors agree well in time and frequency domain Time (sec) 6 Imote2 CV-373 Power spectral density (mg 2 /Hz) $10,000 $500 Imote2 CV Frequency (Hz)

7 Sensor Deployment Cable : 8 Deck : 22 Pylon : 3 Total : 33 In total, 420 channels of sensors Cable : 8 Deck : 26 Pylon : 3 Total : 37 Vibration Sentry 0 Amemometer interfaced with Wind Sentry Nodes underneath deck Nodes on pylons Nodes on pylon top (powered by solar cell) Nodes on cables Nodes on cables (powered by solar cell) Reference Nodes Wind Sentry

8 Vibration measurements 5 Vibrations of the bridge deck, pylons, and cables are captured. Identified modal properties agrees well with FEM analysis acc.(mg) acc.(mg) acc.(mg) Raw acceleration data (Y axis) 5 0 Raw acceleration data (X axis) Raw acceleration data (Z axis) time(sec) Deck vibration 1 st vertical (DV1) : 0.439Hz 0.442Hz 2 nd vertical (DV2) : 0.641Hz 0.647Hz 3 rd vertical (DV3) : 1.025Hz 1.001Hz 4 th vertical (DV4) : 1.404Hz 1.247Hz 5 th vertical (DV5) : 1.569Hz 1.349Hz 6 th vertical (DV6) : 1.837Hz 1.735Hz

9 Full-scale bridge vibration measurements: Main span 570 m 49 nodes along the side walk. Prompt installation Installation 90min, Removal 45min by 3 persons

10 Performance improvement Comparison with past measurements High order modes are identified(upto 17 th mode) page10/15

11 Interdisciplinary and international collaborations antenna Vibration analysis PCB design Sensor, ADC Independent CPU research would take long beyond prototypes Energy Harvesting Network, RF communication Interdisciplinary research involving researchers worldwide accelerate the advance 11

12 Collaboration participants In total, 47 research groups in 8 countries KAIST University of Tokyo UIUC Civil & CS 12

13 Conclusion WSN is considered to allow dense instrumentation on large structures. Key to comprehend their complex behaviors and performances. SW/HW developments on a WSN platform resulted in full-scale bridge vibration monitorings, indicating the merit of WSN dense instrumentation. International/interdisciplinary collaborations are in progress to enhance the technology. 13

14 Vehicle Intelligent Monitoring System (VIMS) 14

15 Vehicle Intelligent Monitoring System (VIMS) 15

16 System components & Presentation of results 16

17 Measurement in Bangkok Date : Total Length : about 304km White Point : IRI 2mm/m 17

18 Train Intelligent Monitoring System (TIMS) 18

19 Train Intelligent Monitoring System (TIMS) DAQ, PC Accelerometer +GPS Track inspection car (East i) Accurate Infrequent measurement TIMS Track condition is estimated from dynamic response of a moving train Frequent measurement 19

20 Data processing 2 Error in localization Before matching Acceleration (m/s 2 ) Acceleration (m/s 2 ) Distance (m) After matching Localization has noticeable error. Repeated measurements on the same track are not directly comparable. Need for Matching Distance (m) 20

21 Change After Local Repair Services Acceleration was measured before and after repair service. Acceleration RMS (m/s 2 ) Repair service Before After 5 out of 12 locations of repair services were detected Vertical acceleration RMS Distance (km) Change in rail condition can be detected. 21

22 Thank you 22