Emerging technology for component life assessment

Operating plant component damage and failure experience is reviewed. Loading conditions such as thermal stratification and striping, turbulent flow and flow-induced vibrations are often found to limit useful life, even though such loadings were typically not considered when the components were designed. High cycle thermal and mechanical fatigue are identified as important damage mechanisms. A new method of correlating fatigue data and extrapolating to the very high cycle regime is described.

The results of environmental degradation testing during the past fifteen years have shown that such effects are much more deleterious than previously assumed. Therefore environmental and aging effects must be taken into account in evaluating the reliability and dependability of components for extended periods of operation. Since most of the available data on environmental effects focus on measured crack growth rates, methods of developing improved fatigue life evaluation methods which include environmental effects on crack growth rates are now being developed.

Fatigue tests on polished specimens are characterized by nominal stress amplitudes over yield, where linear elastic fracture mechanics (da/dn vs. ΔK) methods, such as those used in the ASME Code, are not valid. The small plastic zone corrections used in the Code do not account for the plastic crack-driving energies encountered in low-cycle fatigue testing. J-integral solutions are used herein to evaluate the growth of cracks in these specimens. This approach can be shown to correlate the growth of cracks over the entire range of loading from elastic to grossly plastic conditions in widely different geometries and sizes, including the growth of very short cracks for materials of major interest in pressure vessels and piping. It can be used to correct S-N fatigue life evaluation, curves for known differences in crack growth rates whether they are due to corrosion-assisted fatigue or other variables.

Environmental effects on the crack initiation phase of fatigue failure can be directly incorporated into S-N life evaluation curves. Once the crack propagation effects are included, the resulting improved S-N curves provide a means for plant operators to evaluate the current condition of these components and systems, taking into account the cumulative damage from operating transients and cycles which the plant has experienced. The safe residual life can then be evaluated using the S-N curves to include cumulative damage for the anticipated future period of operation. This plant life evaluation approach is applicable even where in-service inspections are not feasible. It provides a sound quantitative basis for making repair/replacement decisions.