Development Of Mobile Testing And Noncontact Optical Technologies For Dynamic Identification Of Bridges

Bridge construction in China has made remarkable achievements,but bridge collapse accidents occur frequently,causing serious casualties and economic losses.Therefore,how to ensure its safety and longevity is a major national demand.Structural intelligent inspection and health monitoring technologies are expected to find structural defects and hidden dangers in a timely manner.However,structural health monitoring technology can only identify basic modal parameters,which are insufficient to support structural health monitoring.And traditional contact-type sensing technologies are labor-oriented,time-consuming and expansive,which are difficult to achieve rapid measurement and condition assessment of bridges.Therefore,this dissertation focuses on the development of moble testing and noncontact measurement techniques for dynamic parameters identification of bridge structures.The developed techniques are helpful for identifying more structural parameters and improving measurement efficiency.The main research task and innovations are described as follows:(1)Traditional ambient vibration testing methods can only output structural basic modal parameters,which cannot be directly adopted for structural condition assessment.To address this problem,three mobile testing methods including impact vibration test,mass-changing strategy and moving mass technique,were developed for modal flexibility identification.The relationship between unscaled Frequency Response Functions(FRF)estimated from ambient vibration testing and those estimated from impact vibration testing data was firstly investigated(Chapter 2),and then the algorithm for mass scaling factor,modal mass and modal flexibility was developed by using impacting forces(input)and structural responses(output).A mass-changing strategy was further developed in Chapter 3 for bridge rapid testing and flexibility identification by using structural responses only(output).In this chapter,the formula for calculating structural scaling factor was firstly derived based on the first-order sensitivity analysis.Then,the theory of uncertainty quantification of scaling factor and modal flexibility based on Bayesian inference and uncertainty propagation principle was developed.To further improve bridge testing efficiency,a moving vehicle strategy was proposed in Chapter 4 for modal flexibility identification.The relationship between massnormalized scaling factor and time-varying dynamic properties of the coupled system,moving vehicle information was theoretically derived,then the uncertainty quantification theory of modal flexibility was also developed.The proposed three methods can identify more meaningful structural parameters including scaling factor,modal mass and modal flexibility and their confidential intervals.The effectiveness and robustness of the proposed method has been verified by using numerical simulation,experiments of simply-supported beam,and field testing data of Sutong Yangtze River Bridge.It should be noted that structural flexibility can be identified from output-only measurements in mass-changing strategy and moving mass technique.(2)Previous three methods depend on the acceleration measurements by contacttype sensing technologies,noncontact optical measurement-based modal flexibility methods were developed to further improve the measurement efficiency.A vision-based impacting force reconstruction and system identification of footbridge was proposed in Chapter 5.In the proposed method,an optical camera was firstly adopted to acquire the motion image sequences of pedestrians and the pedestrian-induced impacting force was reconstructed by using computer vision technologies,and structural flexibility matrix was identified by combining reconstructed impacting forces and structural responses.A reference-free mobile vision-based measurement method has been proposed in Chapter 6 to further improving testing efficiency and identifying structural flexibility matrix.In the proposed method,vibration images of the entire structure are measured by moving a single camera,from which multipoint dynamic displacements of each setup are calculated by computer vision algorithms and dynamic properties of the structure at several setups can be identified.After that the mode shapes and flexibility matrix of the entire structure can be obtained by integrating basic dynamic properties identified from all measurement setups using the principle of minimum potential energy.The proposed methods can measure impacting forces(input)and structural responses(output)without physical contact,and the effectiveness of the proposed method has been verified by using experiments of simply-supported beam,cantilever beam and footbridge.(3)It is time-consuming for conventional contact-type sensors for cable force estimation and it is very dangerous to install sensors on bridge decks when accident events happen on long-span bridges.Noncontact measurement of long-span bridges were developed for dynamic identification to overcome those issues(Chapter 7).A noncontact cable force estimation method based on the optical imaging technology and unmanned aerial vehicle(UAV)was firstly proposed to improve the testing efficiency for cable tension forces.A line segment detection and matching algorithm for cable displacement extraction was proposed to address the problem of erroneous matching in traditional template matching algorithm.And the combination of relative displacement and the difference of higher modal frequencies of bridge cables was adopted to overcome the problem of UAV hovering.Then,noncontact vortex-induced vibration(VIV)measurement of a long-span suspension bridge was further developed.In this section,noncontact optical sensing technologies were employed to measure multiplepoint displacements of the studied bridge under VIV events;and dynamic properties of the bridge(i.e.natural frequency,damping ratio,mode shapes)and characteristics of the VIV event(i.e.single-mode vibration,dominant vibration mode switch)were identified by analyzing monitoring data.The developed methods have been applied to the noncontact cable force measurement of Nansha suspension bridge and the VIV event measurement of Humen suspension bridge.Results show that the noncontact measurement results have a good agreement with conventional contact-type sensing technologies.The proposed noncontact measurement scheme has good accuracy and greatly improves the measurement efficiency.In conclusion,the developed methods in this dissertation successfully identified structural basic modal parameters and more meaningful structural parameters including scaling factor,scaled FRFs and modal flexibility,and confidential intervals of structural parameters were also quantified.Moreover,the impacting forces(input),structural responses(output),basic dynamic properties and cable tension forces of long-span suspension bridge were measured noncontactly with developed optical techniques.The obtained results in this dissertation are helpful for rapid condition assessment,realibility analysis and long-term performance degredation investigation of bridge structures.