The overall objective of this study is to provide a proven methodology to allow the transfer of ductile fracture initiation properties measured in standard labora tory specimens to large, complex, flawed structures. A significant part of this work involved specifically addressing effects of constraint on transferability under large scale yielding conditions. The approach taken was to quantify constraint effects through micro-mechanical fracture models coupled with finite element generated crack tip stress-strain fields to identify the local condition corresponding to fracture initiation. Detailed finite element models predicted the influence of specimen geometry, loading mode, and material flow properties on the crack tip fields.
The ability of two local, ductile fracture models (the Rice and Tracey void growth model  (VGM) and the stress-modified, critical strain (SMCS) criterion of Mackenzie et al. and Hancock and Cowling [2,3]) to predict fracture initiation were investigated. Predictions were made using experimentally verified, two- and three-dimensional, finite strain, large deformation, finite element analyses. Two, high toughness pressure vessel steels were investigated: A516 Gr70, a ferritic, carbon-manganese mild steel demonstrating high hardening behavior, and HY-80, a martensitic, high strength low alloy (HSLA) steel possessing medium hardening ability. Experimental verification of the ductile frac ture initiation predictions was performed in a variety of crack geometries possessing a range of a/w ratios from 0.15 to 0.70 and experiencing a range of load conditions from three point bending to nearly pure tension. The predicted constraint dependence of global ductile fracture parameters in the two materials is shown.