stress rupture

Creep is about how fast a hot, loaded part slowly stretches; stress rupture is about the end of the story — when it finally breaks. A stress-rupture test loads a specimen at high temperature and simply waits, timing how long it lasts until it snaps. The single answer it gives, the time to failure, is often what a designer most wants to know: how many hours will this part survive at this stress and this temperature?

It is essentially a creep test run all the way to fracture, at higher stresses so failure comes in a testable time. The output is a stress-rupture curve: applied stress plotted against time to rupture (on log axes), usually sloping down — the lower the stress, the longer it lasts. Because service lives (say 100000 hours, over 11 years) are far too long to test directly, engineers run shorter tests at higher temperatures and use a time-temperature parameter — most famously the Larson-Miller parameter, roughly T times (C + log t) — to trade higher temperature for longer time and extrapolate to real service conditions.

Stress rupture is how high-temperature components are actually rated and life-limited: a turbine blade or boiler tube is given an allowable stress that keeps its rupture time comfortably beyond its planned service life, with margin. Two honest cautions come with it. The failure is a creep failure, so it can be preceded by a lot of creep strain (the tertiary stage), and the extrapolation from short, hot tests to long, cooler service is only as trustworthy as the assumption that the same failure mechanism operates in both — which is not always true, so extrapolating too far is a known trap.