The Research On Three Degree-of-freedom Forced Vibration Method For Identification Of Aerodynamic Derivatives And Flutter Mechanism

The difficulties in the field of bridge flutter primarily lies in the exact descriptions of non-steady self-excited forces acting on bridge decks as well as scientific explanations of the mechanism leading to bridge flutter. Being a kind of important parameters describing the aerodynamic performance of bridge decks, aerodynamic derivatives are of vital importance in characterizing the self-excited forces and in working out the flutter mechanism. The identification of aerodynamic derivatives of bridge decks and the flutter analysis based on the aerodynamic derivatives receive a wide attention from bridge wind engineering community due to their dually theoretical and experimental difficulties. A 3-DOFs forced vibration device is first developed in this dissertation to measure the self-excited forces on bridge decks and to identify the aerodynamic derivatives. Then the identification method of the aerodynamic derivatives and the characteristics of self-excited forces are studied in detail. Eventually, a single parameter searching method of 2-D flutter analysis employing a 3-DOFs sectional model and the identified aerodynamic derivatives is proposed in complex domain, and the flutter mechanism is further explored by the 2-D flutter analysis method. The content and research outcome of this thesis is summarized as follows:1. A completely numeral-controlled 3-DOFs forced vibration device is developed to drive the sectional model to oscillate following seven kinds of combination motion of the three DOFs. The motion signal of each DOF is standard sinusoidal wave driven by a machine. And the vibration displacements of the model are measured by the feedback of photoelectrical encoders.2. An experiment setup consisting of composite sectional model with three separate sections is explored for the first time to measure the aerodynamic forces on the sectional model. This setup is capable of measuring synchronously both the forces and pressures acting on the middle dynamometric section, and the measured pressure can be used to obtain the pressure distribution on the model. The self-excited forces for an aerofoil section are measured by force balance method, dynamometric section method with pressure integral technique, respectively, to validate the setup. Then the aerodynamic derivatives are identified through the measured/integrated aerodynamic forces. It is shown that the dynamometric section model with pressure integral technique is of efficacy. 3. A frequency domain method to identify the 18 aerodynamic derivatives by forced vibration tests is studied. The FFT method is further investigated and a CORR algorithm is proposed to identify the aerodynamic derivatives in comparison with those obtained by FFT method. The identification error of frequency-domain method is discussed, and it is found that frequency leakage error in frequency-domain method has no influence on the aerodynamic derivatives from 1-DOF vibration test but will affects the accuracy of aerodynamic derivatives from 3-DOFs vibration test. The correlation between the force and displacement signals in 1-DOF vibration is better than that in 3-DOFs vibration.4. An improved time-domain method to identify the 18 aerodynamic derivatives by forced vibration tests is presented. The velocity signals needed for aerodynamic derivatives identification are obtained by the discrete differentials of the measured displacement signals. The new method circumvents the complicated process in numerical integration of acceleration data in the conventional method to obtain the velocity data and improves the identification precision of the aerodynamic derivatives.5. Several influence parameters of the aerodynamic derivatives of bridge decks are experimentally studied. It is shown that the amplitudes of coupled vibrations have an notable effects on flutter derivatives identification while the effect of frequency of coupled vibrations seems trial. It is also found that flutter derivatives are bi-variate functions in terms of wind velocity and vibration frequency and the flutter derivatives identified by changing the wind velocity while keeping the vibration frequency unchanged deviate appreciably from those identified by changing the vibration frequency while retaining the wind velocity unchanged. Additionally, it is experimentally verified that the appendages on bridge deck will affect the identified flutter derivatives.6. The asymptotic behaviors of the aerodynamic derivatives are derived theoretically by defining the components of self-excited forces with the identified flutter derivatives and investigating the characteristics of flutter derivatives. The energy contributed by each component of the self-excited forces is obtained, and mechanism of bridge flutter is then re-visited from a new viewpoint. The higher-order harmonic characteristics of self-excited forces are also investigated by using the measured forces for a number of cross sections.7. A single parameter searching method of 2-D flutter analysis employing 3-DOFs model and identified aerodynamic derivatives is developed in complex domain. Making use of this method, the equation governing bridge flutter is solved without iteration. The flutter mechanism is explained from a variety of the parameters such as vibration frequencies, damping ratios, phases and energy ratios in the process of the calculation. Finally, the adverse influence of aerodynamic interference between two parallel decks on the flutter stability is investigated through the spring-suspended sectional model tests.This research is financially supported by National Science Foundation of China (No.50478051) and the research fund for Doctoral Program from Ministry of Education, China (No.20040532019).