Karhunen-Loeve展開在 Abaqus 中的實現(xiàn)

打開文本圖片集
中圖分類號:TP391.9;TB115.7 文獻標志碼:B
Application of Karhunen-Loeve expansion in Abaqus
HU Yezhi1, ZHANG Yaqi
(1.DNE Technology Co.,Ltd.,Shanghai 200030,China;2.ALLBRIGHTLaw Ofices(Hefei),Hefei 230001,China)
Abstract:A numerical algorithm for Karhunen-Loeve expansion is proposed based on the theory of numerical analysis. The algorithm is embedded into the computational kernel of Abaqus through a hybrid coding approach using Python and Fortran. The key elements in the development of the program are analyzed,and an error analysis is conducted. The results show that the developed program is feasible, the random field expansion is reasonable,and the numerical algorithm has high precision.The calculated eigenvalues,eigenfunctions,and covariance functions are consistent with the theoretical solutions.
Key Words: random field ; Abaqus; Karhunen-Loeve expansion; covariance function; Fredholm integra equation of the second kind
0 引言
隨機有限單元法(SFEM)是基于概率論對有限單元法(FEM)的擴展補充,其將不確定性分析引入有限元計算中,采用隨機事件描述模型的邊界條件、載荷以及材料參數(shù),其計算結果不再是一個確定的單一數(shù)據(jù),而是符合隨機場理論的一組數(shù)據(jù)的集合,通過概率密度關系能得到一定置信區(qū)間內(nèi)的計算結果[1]
Karhunen-Loeve展開(以下簡稱K-L展開)法是廣泛使用的隨機場離散方法之一,由GHANEM等[2在結構的隨機有限元分析中首次使用,并推導協(xié)方差函數(shù)的特征函數(shù)和特征值需滿足的條件。(剩余6340字)