Comprehensive creep-pore diffusional growth calculations vs. previous approximations

Abstract

Brittle rupture—via nucleation/growth/interconnection of grain-boundary (g.b.) creep pores—is a major cause of failure for materials not-resisting creep (e.g., through precipitation); and, the prediction of the relevant rupture-time with its dependency on stress/temperature/etc. represents a major challenge. Due to mathematical complexity, previous pore-growth calculations contain simplifications with indeterminable errors. Some calculations utilize 1-D (g.b.-vacancy) diffusional models having multi-sized/multi-spaced (‘infinitely’) long-parallel cylinders/pores instead of real-world spheroids (i.e., allowing for diffusion only between closest-neighbors). Other calculations utilize realistic 2-D g.b.-diffusional models but of equi-sized/equi-spaced spheroidal pores (hence ignoring ripening/coalescence); or, of just a couple of configurations of non-uniform/randomly spaced/spheroidal pores (thus not allowing for reliable statistics). Herein, the use of an innovative simulation-technique permits the examination of hundreds of complicated configurations with exceptionally low (w.r.t. the literature) computer-resource-expenditure, leading to proper adjustment/correction of all previous relevant 1-D/2-D works, making their results ‘real-world’ (on major issues, i.e., creep-rupture-time estimations and dependencies on stress/temperature/etc.).