It is very important to correctly model and identify damping in all vibration systems of physical devices that involve damping.
A variety of methods and techniques of damping identification have been developed, most of which can be classified into frequency domain (circle-fitting method, half-power bandwidth method, wavelet transform method, etc.) and time domain (logarithmic decrement method, least squares complex exponential (LSCE) method, Smith least squares method, Hilbert transform method, etc.). Some of the damping identification methods in frequency domain have high accuracy, but the algorithm is very complicated, which restricts its wide application in practice.
Matrix method and modal method are also used for damping identification. In the modal method, modal identification is a must as the damping characteristics are decided by means of modal damping ratio. Modal identification is complicated and it is difficult to ensure the correctness of modal parameters. In the matrix method, which directly identifies the damping matrix by experimentally measured FRF, the imaginary part of experimentally measured FRF is decided for damping identification. This method is highly accurate and it is possible to decide the damping ratio of mechanical structures of lower DOF. However, it is very difficult for all numbers of DOF as mechanical structures are of high DOF in practice.
Kim Won Ju, a researcher at the Faculty of Mechanical Science and Technology, has proposed a method for correctly and simply deciding the damping characteristics of the whole system using the experimentally measured FRF and the FE model. Using the experimentally measured frequency response function matrix, he built an experimental model and combined it with the FE model to enlarge the experimental identity data to the whole system.
Comparing his method with preceding ones through simulations and error analysis, he verified its precision and effectiveness.
© 2021 KumChaek University of Technology