质量时变系统的动载荷识别

研究质量时变系统的动态载荷识别技术,引入广义模态概念,在初瞬时的模态空间中利用模态叠加法得到部分解耦的动力学方程,结合考虑质量变化效应的虚拟变形法和Duhamel积分,构建反演动态力的2种分析途径.一种是基于加速度反演的连锁递推格式,另一种是基于位移和加速度反演的正则化求解格式.分析2个动载荷作用下的变质量多自由度系统,利用Euler中点辛差分方法求得系统的响应,将其叠加随机噪声以模拟测点响应,然后,应用2种模态力反求格式反演动载荷.研究结果表明,连锁递推格式计算效率高,但对加速度测量噪声敏感;正则化求解格式的稳健性较高.

Identification of Dynamic Loads in Mass variable System

The technique of the load identification of mass variable system was studied. The generalized modal concept applied on time varying system was introduced, and the method of modal superposition was used to give dynamic equations with partially decoupled in initial modal space defined at initial instant. Based on both virtual distortion method (VDM) accounting for the effect of mass change and the Duhamel integration in initial modal space, two approaches to identify modal forces were constructed: one is the recursive calculation pattern formulated on the basis of modal acceleration, and the other is the regularized solution by the direct discretization of Duhamel integration with modal displacement and acceleration. A mass variable system with 16 degrees of freedom subjected to two external loads was analyzed. The responses were calculated by applying Euler midpoint symplectic difference method and added with random noise to simulate measured signals. Two dynamic loads were identified by two inversion approaches of modal forces. The comparison of results has indicated that regularized solution format has a higher numerical robust than the recursive calculation pattern, but the latter is sensitive to the noise of measurement.