Development of Experimental Identification Techniques in Time Domain for Boundary Conditions and Joint Conditions of structures

确定机器和结构的边界条件和结合部条件是实验识别领域中最困难的问题之一。为了预测它们的行为，有限元法等数值方法被广泛使用。为了得到准确的结果，边界条件和关节条件的准确描述是必不可少的。因此，发展边界条件和联合条件的识别技术是十分重要的。在这项研究中，我们的目标是开发通用的识别技术。作为第一步，我们在2001年开发了一种计算梁的非线性边界条件的技术。这种技术是基于谐波平衡，因此使用稳态响应周期激励。通过数值模拟和实验验证了该方法的适用性。然后我们推广了该技术，使其可以应用于二维问题。作为典型实例，我们开发了一种适用于非轴对称圆板的方法，并通过数值模拟验证了该方法的适用性。然后，作为我们的最终目的，我们在2002年开发了一种使用时间序列数据的技术。与上述技术相比，时间序列数据的使用减少了对激励类型以及响应数据的限制。作为一个典型的例子，我们提出了一种适用于梁的技术。通过数值模拟验证了该方法的适用性。

One of the most difficult problems in the field of experimental indentification is to determine the boundary conditions and joint conditions of machines and structures. To predict their behavior, numerical methods such as the finite element method are widely used. To get accurate results, accurate description of the boundary conditions and joint conditions is essential. Thus development of identification techniques of boundary conditions and joint conditions is of great importance.In this study, we aimed to develop general-purpose identification techniques. As a first step, we developed in 2001, a technique for nonlinear boundary conditions of a beam. This technique is based on the harmonic balance, and hence uses steady-state response to a periodic excitation. We confirmed the applicability of the proposed technique by numerical simulation and experiment. Then we generalized the technique so that it can be applied to two-dimensional problems. We developed, as a typical example, a technique applicable to a non-axisymmetric circular plate, and confirmed the applicability by numerical simulation. Then, as our final purpose, we developed in 2002, a technique that uses time series data. Use of time series data reduces limitations on the type of excitations as well as the responses data, as compared with the above technique. As a typical example, we proposed a technique that is applicable to a beam. We confirmed the applicability by numerical simulation.