Estimation of mixed-mode fracture parameters by gene expression programming

Abstract

The linear elastic fracture phenomenon has been characterized with stress intensity factors (SIFs). In this study a general function is obtained in order to predict the fracture parameters. Numerical calculation of the SIFs in a mixed-mode condition is a cumbersome task. In this research, more than 6800 numerical analyses using extended finite element method are conducted to simulate the fracture problem. States are considered for a plate with an arbitrary edge or center crack. Mixed mode SIFs were calculated using of interaction integral. Then, Gene Expression Programming (GEP) method is utilized to extraction of a function. Results show acceptable correlations between numerical calculations and genetic programming functions. R-square (R2) values are in a range of 0.91 to 0.96 that guarantees the accuracy of the inferred functions.