Modal Parameter Identification Of Aeroelastic Models Of Long-span Bridges

Wind tunnel experiments of full aeroelastic models are becoming an importanttool for investigating wind-induced performance of long-span bridges. In order toobtain the epxperimental results representative of full-scale values, the similarity ofthe modal properties as well as the similiarty of geometrical shapes between themodel and the prototype should be followed. Due to small model scales, the designprocess of aerleastic models must introduce some simplications and the modelfabrication process may also produce some inaccuracy. In order to assure the fidelityof the model, modal parameter identification from aeroelastic models is of paramountsignificance. The major paramters describing a model in frequency domain are modalfrequencies, mode shapes, modal damping ratios as well as modal mass; as long as themass and mode shape are consistent between model and prototype, the modal masswill also be consistent. As such, this article purses the identification of modalfrequencies, mode shapes, and damping ratios for the aeroelastic models by using thefrequency domain decomposition (FDD) and data-driven stochastic subspaceidentification (SSI) available in the commercial package ARTeMIS. The main work ofthe thesis is summarized as follows.(1) After a review of existing literature, the theories and algorithm process ofmodal identification methods including FDD and SSI are described. Three differentrealizations of SSI algorithms due to the different weights assigned to projectionmatrix, namely UPC, PC and CVA, are discussed.(2) Making use of ARTeMIS, modal identification of a simply beam bridge andLishui suspension bridge are carried out from different incentive condition simulatedacceleration data polluated with noise. The results indicated that the modalfrequencies and mode shapes are identified with reasonable accuracy. However theidentification results of damping ratios are not satificatory whose values dependes onappropriate choice of some paramters.(3) Detailed introduction of aeroelastic models of Lshui Bridge in Hunanprovince, as well as the ambient vibration test. It is shown that the modalfrequencies and mode shapes are of reasonable accuracy. The identified modaldamping are large.