Stress Analysis of Layered Elastic Solids With Cracks Using the Fast Fourier Transform and Conjugate Gradient Techniques

2001 ◽  
Vol 68 (5) ◽  
pp. 708-714 ◽  
Author(s):  
I. A. Polonsky ◽  
L. M. Keer

The fast Fourier transform (FFT) technique has recently been applied to stress analyses of layered elastic solids, with a great deal of success. However, the existing FFT-based methods are limited to intact solids. This paper explores the possibility of using FFT for stress analyses of layered elastic solids containing cracks. A new numerical approach is developed by combining three-dimensional FFT with the theory of periodic eigenstrain and the conjugate gradient method. The new method is primarily designed for analyzing complex three-dimensional crack patterns in layered solids, such as those produced in thin protective coatings by roughness-induced contact stresses. The method should be particularly advantageous for studying crack propagation in coatings, as it does not require remeshing when the crack shape changes. Numerical examples illustrating advantages as well as limitations of the method are presented. Some unexpected results that were obtained for multiple cracks in a thin coating are discussed.

Author(s):  
Zhanjiang Wang ◽  
Xiaoqing Jin ◽  
Shuangbiao Liu ◽  
Leon M. Keer ◽  
Jian Cao ◽  
...  

This paper presents a new method of contact plasticity analysis based on Galerkin vectors to solve the eigenstresses due to eigenstrain. The influence coefficients relating eigenstrains to eigenstresses thus can be divided into four terms the one due to the eigenstrains in the full space, others due to the mirrored eigenstrains in the mirror half space. Each term can be solved fast and efficient by using the three-dimensional discrete convolution and fast Fourier transform (DC-FFT) or the three-dimensional discrete correlation and fast Fourier transform (DCR-FFT). The new method is used to analyze the contact plastic residual stresses in half space.


Sign in / Sign up

Export Citation Format

Share Document