动态有限元模型修正的一种新方法

目前采用实验模态数据修正动态有限元模型的方法,把原带状稀疏的矩阵修正为满值矩阵,这是一个较大的缺陷。提出了一种新的修正方法,该方法借助于数学上的拉直运算,把需修正的变量分离出来直接对其进行修正运算,这样就可以保证质量矩阵与刚度矩阵带状稀疏的特点。同时,为了防止修正后质量与刚度矩阵变化过大,本文对位于质量矩阵和刚度矩阵带状内变量的变化范围进行了限制。本文算法不仅计算量小,而且精度较高。最后给出了计算实例。算例表明该方法具有较高的修正精度。

Improved Straightened Algorithm Using Experimental Modal Data to Modify Calculating Model

Great difference exists between the results by the original model and the modified results by adopting the straightened algorithm in which experimental modal data are used to modify calculatin model. In the paper, an improved straightened algorithm is introduced to make up the defect, and the modified model can be acquired by aid of limiting the range of all confined variables, and of optimization as well as iteration. Meanwhile the fact that the values of partial elements within taenioid matrix are equal to zero is also considered. The calculated results by using improved algorithm are close to that by the original model and in good accordance with the actual case and high precision can be assured. Finally, a numerical example is presented.