Numerical Verification of Bridge Screening Technology based on Vehicle Vibration

Transcription

1 Numerical Verification of Bridge Screening Technology based on Vehicle Vibration University of Tsukuba Kyosuke Yamamoto Assistant Professor

2 Background: Bridge Inspections based on Visual Check by Veteran Engineers

3 Social Matter: The Training of Engineers requires several years, though there is a Large Number of Aging bridges Which bridge should I check? To implement the priorities of bridges we need to know the soundness before the inspections!?

4 Solutions: Bridge Screening (Roughly Quick Inspection) enables Strategic Allocation of Resources! Bridge Screening based on vehicle responses Detailed Inspection Sensors on Vehicle Intact Intact Damage

5 Technical issue: How to estimate the bridge status only by using vehicle vibrations? 1 To measure the vibrations of the passing vehicle 2To transform them to vibrations at the fixed points First Mode 3To estimate the mode shapes of the bridge Second Mode Spatial Singular Mode Angle: θ = tan 1 A 21 A 11 = tan 1 A 12 A 22

6 The Purpose: To examine the applicability of SSMA to the Bridge Screening Bridge Vibration y t = Aq t Bridge Vibrations measured at Moving Positions N t 1 y t = Aq t Vehicle Vibrations (1) Bridge vibrations (2) Road Roughness (3) Vehicle Body vibrations N t 1 z t = Aσ t

7 The Proposed Method: Application of SVD method to the spatially corrected vibration data Measured Vehicle Vibrations z t SVD: Singular Value Decomposition N t 1 z t x t = A s t Spatially Corrected Vibrations at assumed fixed points Estimated Mode Shapes Estimated Basis Coordinates SVD assumes their non-correlation =

8 The Methodology: Experimental & Numerical Verification Data Acquisition Numerical Simulation Field Experiment Laboratory Experiment Application of the proposed method

9 The Numerical Simulation: RBSM Vehicle & FEM Bridge t (=2kN) Vehicle Model 4m span Steel Truss Bridge Displacement[mm] Road Unevenness Position[m]

10 The Procedure to calculate SSMA: Spatial Correction & Mode Decomposition Acceleration[m/s 2 ] Mode shape 1 X Y Acceleration[m/s 2 ] Wheel Position[m] L/3 2L/3 L Position [m] Y First Mode Second Mode X

11 The Results of Numerical Simulation: SSMA Applicability to the damage detection when varying velocity & road roughness SSMA[deg] Intact Damaged V elocity[m/s] Smooth Road SSMA[deg] Intact Damaged V elocity[m/s] Rough Road

12 The Robustness: It is found that SSMA has the high sensitivity to a local bridge damage 割合 割合 Frequency Frequency SSMA[ ] Smooth Road 割合割合 SSMA[ ] Rough Road

13 The Field Experiment: SSMA Applicability to the damage detection on the actual steel truss bridge 1) ボルト固定 2) 切断 3) ボルト除荷 車上計測 On-going Measuring 損傷検知 Damage Detection 11mm 25ton Truck 25t トラック 546mm 546mm 546mm 546mm 546mm 546mm 546mm 546mm 546mm 546mm 546mm 546mm 65m span 6552mm 前軸 後軸 総重量 Front Rear Total INTACT 健全時 DAMAGE 損傷時 (kn) (kn) (kn) 1(km/h) 2(km/h) 4(km/h) 1(km/h) 2(km/h) 4(km/h) First Day 1 日目 日目 Second Day First First 1 回目 1 回目 1 回目 2 回目 2,3 回目 Second First Second Third First Second Third First Second Third First Second Third 1~3 回目 1~3 回目 1~3 回目

14 SSMA The Results of Field Experiment: High sensitivity to the damage detection Low Accuracy to the damage identification 〇 INTACT 〇 DAMAGE Because the mode shape is independent from the vehicle speed, it means the low estimation accuracy of SSMA. Speed (km/h) The high sensitivity to damage because the angle changes distinctively after damage.

15 1 6 5 The Lab. Experiment: Statistical Validity of SSMA should be examined for Engineering application ピン支承 CH2 2 CH1 CH3 CH4 ピン支承 Damage Section Sensor センサーレール Rail Truck レール除去部 Removed Rail アプローチ部 進行方向 アプローチ部 ワイヤー

16 The result of the Lab. Experiment: SSMA probability density distribution statistically changes after the bridge damage.95.9 SSMA [rad] SSMA [rad] Probability Run speed [rpm] Speed [rpm] SSMA [rad]

17 Discussion: Why does SSMA shows the high sensitivity to the bridge damage? x t = A s t Probability Histogram Normal distribution.1 Intact Probability Histogram Normal distribution.1 Damage Kurtosis of reference coordinate

18 Discussion: Why does SSMA shows the high sensitivity to the bridge damage? x t = A s t Noise INTACT Damage DAMAGE INTACT Damage DAMAGE INTACT Noise Damage & SVD error Mode Shape What SVD can estimate is the component of non-correlation close to the signals Basis Coordinate DAMAGE

19 Conclusion: 1. SSMA is the bridge mode shape estimated only by using the vehicle data. 2. The applicability and statistical validity is verified numerically and experimentally. 3. It is applicable to the damage detection because of its high sensitivity. 4. It is still difficult to apply SSMA to the damage identification, because of its low estimation accuracy

20 University of Tsukuba 1-1-1, Ten-Nou-Dai, Tsukuba, Ibaraki, JAPAN