首页> 外文学位 >Improved experimental impedance modeling techniques for updating finite element models.
【24h】

Improved experimental impedance modeling techniques for updating finite element models.

机译:改进的实验阻抗建模技术,用于更新有限元模型。

获取原文
获取原文并翻译 | 示例

摘要

In the design and optimization of a structure or machine, an accurate finite element model is often necessary to predict the structure's noise and vibration behavior. This dissertation investigates some important aspects of experimental impedance modeling techniques and proposes the improved techniques for updating finite element models.; An SVD/QR based model error indicator function is proposed to improve the error localizing procedure. A singular value decomposition is first used to select the meaningful portion of a least squares problem data matrix. Then, a QR permutation pivoting decomposition is applied on the selected matrix to find out the parameters that correspond to the dominant model errors. This method is efficient since both SVD and QR decomposition methods have a good ability to condense large data matrices and is effective since the meaningless portion of the data matrix is discarded.; The experimentally measured data contains damping, while damping is generally not available in the finite element model. An improved finite element model updating procedure is proposed to handle this inconsistency. In the improved algorithm, the dominating elements in a finite element damping matrix are eliminated by matrix multiplication operations. The improved procedure is used successfully for heavily damped structures. Three strategic methods are proposed for three specific application cases respectively.; Large differences exist among the updated results from different updating methods. This is because of insufficient information about the test structure in the measured data. A Perturbed Boundary Condition (PBC) updating procedure is proposed which utilizes the data from more than one structural configuration. The use of properly selected PBC configurations generally leads to more accurate updated models and so provides better predictions of the structural dynamic properties.; Numerous simulations as well as an experimental examples are used to demonstrate the concepts evaluated in the course of this research. A combined application of all of the proposed techniques are made to update the finite element model of an H-frame structure, emphasizing on the rotational stiffness properties of the joints between box beams. Practical considerations in applying the new methods are addressed.
机译:在结构或机器的设计和优化中,通常需要精确的有限元模型来预测结构的噪声和振动行为。本文研究了实验阻抗建模技术的一些重要方面,并提出了改进的有限元模型更新技术。提出了一种基于SVD / QR的模型误差指标功能,以改善误差定位过程。首先使用奇异值分解来选择最小二乘问题数据矩阵的有意义部分。然后,将QR置换透视分解应用于所选矩阵,以找出与主要模型误差相对应的参数。该方法是有效的,因为SVD和QR分解方法都具有很好的压缩大型数据矩阵的能力,并且由于丢弃了数据矩阵的无意义部分而非常有效。实验测量的数据包含阻尼,而在有限元模型中通常不提供阻尼。提出了一种改进的有限元模型更新程序来处理这种矛盾。在改进的算法中,通过矩阵乘法运算消除了有限元阻尼矩阵中的主导元素。改进的方法已成功用于高阻尼结构。针对三种具体的应用案例分别提出了三种策略方法。不同更新方法的更新结果之间存在很大差异。这是因为测量数据中有关测试结构的信息不足。提出了一种扰动边界条件(PBC)更新过程,该过程利用了来自多个结构配置的数据。使用正确选择的PBC配置通常会导致更准确的更新模型,因此可以更好地预测结构动力学特性。大量的仿真和实验示例用于证明在研究过程中评估的概念。结合所有提出的技术来更新H框架结构的有限元模型,并着重强调箱形梁之间节点的旋转刚度特性。解决了应用新方法的实际考虑。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号