The prediction and verification of ductile crack growth in small test specimens

The ability to predict ductile crack growth in small samples has been of interest to the fracture mechanics community for decades. Previous experimental and theoretical work in the field of elastic-plastic fracture mechanics has enhanced the understanding of decreasing specimen size effects on crack growth predictability. This thesis furthers the understanding of specimen size limitations by experimentally testing three different sized (0.394T, 0.2T, and 0.1T) A516-70 steel three point bend (TPB) samples and modeling them using a computational cell methodology in a finite element analysis (FEA) code called Warp3D. The findings suggest that ductile crack growth prediction may be feasible in experimental samples as small as 0.2T, but not in the comparable FEA model. It is demonstrated, however, that a 0.394T TPB FEA model is capable of predicting crack growth.