Analysis of dynamical processes under consideration of polymorphic uncertainty
Analysis of dynamical processes under consideration of polymorphic uncertainty
复制标题
DOI：
10.1016/j.strusafe.2014.09.003
复制
发表时间：
2015
影响因子：
5.8
通讯作者：
W. Graf;Marco Götz;M. Kaliske
中科院分区：
文献类型：
--
作者：
W. Graf;Marco Götz;M. Kaliske
结构的数值分析一方面需要一个现实的非线性结构模型和相关的计算算法，另一方面需要考虑数据的不确定性。不确定性源于结构参数的变化、对先前激励的不完全了解、数值模型本身的质量和/或关于模型中参数的信息(数据)的缺失。多态不确定性模型的应用使得在不确定参数中考虑多个不确定性特征成为可能。这些模型允许同时考虑可变性、印象和不完全性，并能够对每个不确定性特征对结果的影响进行单独评估。在这些模型的帮助下，结构工程师可以评估改进能力，并采取进一步措施来减少不确定性，例如不完整性。本文提出了两种考虑不确定性的时变结构分析方法。这两种方法在不确定度分析和时间步长积分方面各不相同，可用于现有的确定性数值/解析解。给出的算例验证了两种方法的适用性。
The numerical analysis of structures requires, on the one hand, a realistic nonlinear structural model and related computational algorithms and on the other hand, the consideration of data uncertainty. The uncertainty results from variations in structural parameters, incomplete knowledge about previous excitations, the quality of the numerical model itself and/or missing information (data) about parameters in the model. The consideration of more than one uncertainty characteristic within uncertain parameters is possible by the application of polymorphic uncertainty models. These models allow taking variability, impression and incompleteness simultaneously into account and enable a separated evaluation of the influence of each uncertainty characteristic on the results. With the aid of these models, structural engineers can assess the improvement capacity and take further steps to reduce the uncertainty, e.g. for incompleteness. In this contribution, two methods for time-dependent structural analysis under consideration of uncertainty are formulated. The methods differ from each other with respect to uncertainty analysis and time step integration and can be used for existing deterministic numerical/analytical solutions. The proposed examples demonstrate the applicability of both methods.