Font Size: a A A

Prediction Methods On Vortex-induced Vibration Amplitudes And Sotf Galloping Responses Of Flexible Birdges

Posted on:2014-01-10Degree:DoctorType:Dissertation
Country:ChinaCandidate:S ZhouFull Text:PDF
GTID:1262330428966776Subject:Civil engineering
Abstract/Summary:PDF Full Text Request
Flexible girders of large span bridges are prone to suffering from vortex-induced vibrations (VIV), some practical engineering cases have beenrecorded and reported. The acceleration amplitudes of high mode responsesare increasing by square times of the vibration frequencies, which bring morerisk for the safety behaviors of bridges. Similarly, the slender rectangularcylinders of large span steel trussed arch bridges are frequently suffering fromwind induced vibrations, a kind of soft galloping response produced by thecombined effects of vortex-induced vibration (VIV) and galloping, theamplitude response are continuously increasing with the increasedoncoming wind velocities until the whole structure collapse, which areassumed to be one of the key problem of design process. Referring to the twomain engineering vibration backgrounds of flexible girders and slenderrectangular cylinders of large span bridges, the researches of estimation ontheir amplitude response are performed in the present work, the specificinvestigations may be concluded as following:(1) A slender rectangular steel tube is selected as prototype cylinder toperform the wind tunnel tests, the vortex-induced vibrations and gallopingresponses of fundamental mode are separately measured at different anti-windstatus. According to the dynamic parameters of prototype cylinder, the sectionmodels are designed by the principle of one to one, which means the prototypecylinder and section model are sharing the same c ross section, mass, dampingand natural frequencies. The comparative wind tunnel tests between prototypecylinder and section model are conducted to investigate the3D effects on VIVamplitude and soft galloping response. The test results indicate that the VIVamplitude response of prototype cylinder reached at a higher level thansection model, with the ratio of1.2; while the soft galloping response areobserved to be almost the same.(2) A multi-points flexibly supported aeroelastic model isdesigned to study the high mode behaviors, the high mode VIV responses aremeasured in the wind tunnel tests, the dynamic parameters and VIV amplituderesponses are comparatively analyzed. The section models are individuallydesigned according to the dynamic parameters of5t h,6t hand7t hmode of aeroelastic model by the principle of one to one, the wind tunnel tests resultsbetween aeroelastic model and section models are selected to study the3Deffects on VIV amplitudes. The final comparisons show that t he maximumVIV amplitude of5t h,6t hand7t hmode of aeroelastic model are higher than thecomparatively section models, the ratios are ranging from1.25to1.32.(3) According to the wind tunnel tests results of multi-points flexiblysupported aeroelasti c model, the investigations on correlation length areimplemented by forced vibration wind tunnel tests, the effects of oscillationamplitude, mode shape and different districts of VIV lock-in on aerodynamiccorrelation length are separately studied. The t ests results indicate that theoscillation amplitude and mode orders have positive impacts on aerodynamiccorrelation length; referring to the effect of the VIV lock-in, the correlationlength at onset point, peak point and ending point of VIV lock-in present thegenerally decreasing trend.(4) The CFD based numerical simulations are performed to investigatetwo complicated aerodynamic phenomenon, the one is soft galloping responsewhich are frequently observed in wind tunnel tests, the other one is the tw oseparate VIV lock-in phenomenon of the same2D section model system. Theresults of numerical simulations have good agreements with wind tunnel tests.The wake vortices of the two cases are identified by post-process of CFD, thebehaviors of time serials and frequencies are investigated, the relationship ofphase angle between aerodynamic force and displacement, structuraldisplacement responses and aerodynamic force on wake vortices mode areseparately studied.(5) On the basis of2D Tamura wake oscill ator VIV model, the effects of3D mode shape and spanwise correlation length are taken into account, themodified3D VIV model is proposed in the present work. The3D mathematicalmodel is testified by the wind tunnel tests results of high mode responses o fmulti-points flexibly supported aeroelastic model, the possibilities ofapplication of this3D model on real large span bridges’ VIV amplitudeestimations are discussed.(6) The Runge-Kutta based numerical simulations are implemented tostudy the soft galloping response, numerical simulation results are coincidingwith the wind tunnel tests results. The critical parameter on amplituderesponse of soft galloping is assumed to be the aspect ratio of cross section through comparative calculations. After the mathematical fitting on largeamount of measured data between normalized wind velocities anddimensionless amplitudes of soft galloping, an empirical formula for theamplitude estimation of soft galloping of slender rectangular cylinders isproposed in the present work.
Keywords/Search Tags:Bridges, Vortex-Induced Vibration, Galloping, AmplitudeEstimation, Mode Shape, Spanwise Correlation, Wind Tunnel Tests, Numerical Simulations
PDF Full Text Request
Related items