周波数応答関数を用いた質量・剛性・減衰特性の実用的同定手法

抄録

Vibration reduction is important to improve ride quality of railway vehicles. Damping is one of the essential characteristics of vibration phenomenon and therefore estimating and increasing damping performance of car body is considered as a significant problem. This study is aiming to identify mass, stiffness and damping matrices (called characteristic matrices in this study) from frequency response function which is obtained by some realistic excitation testing. In a previous research, Chen showed a method which estimated characteristic matrices using a response result of vibrating all particles of a kinetic model. However, it’s not realistic to apply this method in railway vehicles because it needs driving point response measurement for all reference points. In this report, we developed a procedure to estimate those matrices from frequency response functions obtained by single driving point. And we validated the proposed procedure numerically by using simple four degrees of freedom kinetic models. Since the identification procedure uses complex natural modal properties of the objective system, we firstly checked the influence by the accuracy in modal property estimation. Then we also investigated the influence of the noise contamination in measured data. It has been found that accurate modal estimation is essentially important to obtain characteristic matrices successfully. And this procedure is relatively robust against noise contamination.