Font Size: a A A

A Study On Damage Identification For Concrete Filled Steel Tubulararch-beam Composite Bridge

Posted on:2015-01-15Degree:DoctorType:Dissertation
Country:ChinaCandidate:J L HuFull Text:PDF
GTID:1262330422481508Subject:Bridge and tunnel project
Abstract/Summary:PDF Full Text Request
Concrete filled steel tubular (CFST) arch bridge is the one whose main load-bearingcomponent is arch rib that is composed of steel tubes filled with concrete. Due to its excellentperformance such as superior capacity, good plasticity and toughness, high fire-resistance andimpact-resistance, easy construction and remarkable economical efficiency, CFST arch bridgehas been widely used since it first appeared in China in1990. However, the development ofits technical such as design theory, computational methods, codes establishment fall behind itsconstruction speed. Thus, there are some certain defects in the design and construction ofCFST arch bridges in the early days. With the long time operation and environmental effect,these bridge diseases led by these defects have been appeared increasingly, which seriouslyaffect the operational safety of these bridges. To determine the bridge structural damageaccurately, and then assess security situation, put forward some practical response measures,ensure the operational safety of these bridges, this dissertation took the CFST arch-beamcomposite bridge-Dukeng Bridge as the background, some issues about the damageidentification of this type of bridge were studied as follows:1) model updating of structuralfinite element (FE);2) optimization of sensors placement and Structural modal expansion;3)damage mechanism, damage index and damage identification on CFST arch-beam compositebridge were investigated;4) and structural damage extent assessment and damaged structuresafety condition evaluation. The main content of this dissertation are shown as follows:(1) The structural FE model updating. Kriging model is presented for structural FEmodel updating in this dissertation, in which the Kriging model parameters is optimized withgenetic algorithm (GA). Taking the design parameter variables and FE structural analysisfrequency as the training sample to train Kriging model, taking the measured structuralfrequency as input sample, then the optimized design parameter variables are outputted.Compared with neural network (NN) method and GA method, the FE model updating methodbased on Kriging model will obtain the higher accuracy of updated model but using a littleamount of training samples, and less time consumed.(2) Sensors placement optimization. A new approach which combine the improvedsingular value decomposition and modal strain energy (MSE) was proposed in thisdissertation. Improved SVD can get the maximum linearly independent combination ofsensors measuring points. Node MSE characterizes the activity of the node. Exponentialfunction was taken as the belief function to calculate the belief of candidate measuring point.And then measuring points belief and node MSE were fused. The candidate measuring points which have greater targets were set as formal measuring points. The measuring pointplacement scheme obtained by this method can get much more modal information with highersignal to noise ratio (SNR).(3) Structural modal expansion analysis. This dissertation introduced three methods ofstructural vibration modal expansion, such as the Kidder dynamic modal expansion method,system equivalent reduction expansion process (SEREP) method and the improved reducedsystem (IRS) method. Implement the modal expansion of a steel truss structure and main girdof Dukeng Bridge by three modal expansion methods. The applicability and accuracy of threemethods was analyzed according to modal expansion results. Modal expansion results showthat all three methods proposed above have high accuracy to simple structures like the trussand SEREP method shows the highest accuracy. However, the accuracy to complex structureslike Dukeng bridge is unsatisfactory. The three modal expansion methods have limitation tocomplex structures real application(4) Damage mechanism for CFST arch-beam composite bridge. CFST arch-beamcomposite bridge was dispersed into3main components in this dissertation, which wereCFST arch rib, pre-stressed concrete main girder, and suspenders system. The mechanisms ofCFST arch rib separation, cracks in pre-stressed concrete main girder and suspenderscorrosion were analyzed. These three damage mentioned above were considered to be mostprone to occur. Especially, this dissertation derived the critical condition of axial pressure, thetemperature difference and concrete shrinkage which lead to CFST arch rib separation.Besides, solid FE model was established to test the correctness of CFST componentseparation critical condition that derived.(5) CFST arch-beam composite bridge damage index. Several common static/dynamicstructural damage indices and damage identification methods were compared with each otherin this dissertation. The author proposed different type of component should apply appropriatedamage index for component characteristics. Wave equation of arc arch was derived in thisdissertation. Wavelet packet was applied to decompose the stress wave signal. And energyratio variation deviation (ERVD) of wavelet packet energy spectrum was taken as the damageindex of CFST arch rib. Strain mode difference was taken as the damage index of pre-stressedconcrete girder and the effective area ratio was defined as recommended value of sensorsplacement density. Difference of internal forces were taken as the damage index of suspenderssystem and NN was applied to identify suspenders damage.(6) CFST arch-beam composite bridge damage extent assessment. A damage extentassessment model of CFST arch-beam composite bridge was established in this dissertation. Fuzzy theory was used to quantify the various factors, then the membership of each factorwas calculated to grade structural damage extent. An actual tested CFST arch bridge damageextent assessment was demonstrated in this dissertation.(7) Safety condition evaluation for damaged structure. An improved response surfacemethod with neural network was applied to structural ultimate bearing capacity reliabilityanalysis. Effective ultimate bearing capacity ratio based on reliability index was defined toevaluate the damaged structure safety condition. Dukeng Bridge was taken as an example todemonstrate the process of safety condition evaluation.
Keywords/Search Tags:CFST arch-beam composite bridge, Model updating, Kriging model, Data fusion, Improved SVD algorithm, Stress wave, Wavelet packet decomposition, Membership, Reliability, Effective ultimate bearing capacity ratio
PDF Full Text Request
Related items