Abstract
The existing component lifetime design and evaluation codes for cracking in engineering structures (e.g., the ASME Boiler and Pressure Vessel Code, Sections III & XI) are empirically-based rules for tolerance to cyclic loading. As such, they to not account for time dependent crack advance or for the continuum in water and material chemistry. Alternative approaches based on fundamental models of environmental cracking have quantified with reasonable accuracy the effects of loading, environment and material variables for metallurgically and chemically "long" cracks. While directly applicable to life evaluation of cracked components, the use of these models for design and overall prediction of component lifetime requires understanding of crack "initiation" (i.e., pre-detection) behavior, where the time required to develop a "long" crack can be a substantial portion of the component lifetime. The delineation of "initiation" invariably depends on semantics and crack detection sensitivity, although a reasonable definition can be based on short crack behavior, i.e., the depth or time required to develop an occluded crack chemistry and/or "long" crack mechanics. Short vs. long crack chemistry can be understood in terms of a critical aspect ratio for enhanced crack tip chemistry; this has been the primary focus of the experimental work to date and can be used, for example, to interpret the effects of creviced geometries. Short vs. long crack mechanics can be understood in terms of their effects on crack tip strain rate, a fundamental parameter in environmentally assisted cracking.
Our objective is to refine the concept of crack "initiation" by examining the crack growth rate behavior in a crack length regime which is shorter than commonly studied. Ongoing experiments are designed to study the behavioral transition from incubation, to "short", to "long" cracks by observing changes in crack growth rate under controlled conditions of constant crack tip chemistry. These studies show that in short cracks, growth rates are limited to ≈≤50 µm and are retarded compared to long cracks. Also, analyses are presented based on ("long") crack growth rate modeling to predict component lifetimes; this approach is based on crack advance from a small intrinsic defect, with comparisons between predicted and actual component lifetime for stainless steel piping from numerous BWR plants. The assumptions and conclusions of the analyses, particularly in terms of the intrinsic defect size, are supported by the experimental data.
