Creep Rupture Fundamentals
Creep rupture — the time-dependent fracture of materials under sustained load at elevated temperature, the physical mechanisms, the stages leading to rupture, and the relationship to the creep curve.
What is creep rupture?
Creep rupture is the fracture of a material under a sustained load at elevated temperature. It is the final event in the creep process — the accumulated creep damage in the tertiary stage reaches a critical level and the material fractures. Unlike ductile fracture at room temperature (which occurs rapidly after yielding), creep rupture occurs slowly over time — the component may operate for thousands of hours before fracturing. The rupture life depends on the stress, the temperature and the material. Creep rupture is the primary life-limiting mechanism for many high-temperature components, including pressure vessels, piping, turbine blades and heat exchangers.
Physical mechanisms
Creep rupture is driven by the accumulation of microstructural damage during tertiary creep. The dominant mechanisms are void nucleation and growth at grain boundaries, and grain boundary sliding. At high temperature, grain boundaries are weaker than the grain interiors — they slide and cavitate under stress. Voids nucleate at grain boundary triple points, particles and serrations. The voids grow by diffusion and by creep deformation of the surrounding material. As the voids grow and coalesce, they form microcracks, which link to form a macrocrack, leading to fracture. The fracture is often intergranular (along grain boundaries), in contrast to the transgranular fracture typical of room-temperature ductile failure.
Stages leading to rupture
The creep rupture process follows the three stages of the creep curve. In primary creep, the material strain-hardens and the damage is negligible. In secondary creep, the damage begins to nucleate (voids form at grain boundaries) but the damage rate is low and the damage is distributed. In tertiary creep, the damage accelerates — voids grow and coalesce, the effective load-bearing area decreases, the true stress increases, and the creep rate accelerates. The acceleration is self-reinforcing until the remaining ligament fractures. The rupture life is the total time from load application to fracture — the sum of the primary, secondary and tertiary durations.
Stress and temperature dependence
The creep rupture life decreases with increasing stress and increasing temperature. At higher stress, the creep rate is higher (Norton law), the damage accumulates faster, and the rupture life is shorter. At higher temperature, the creep rate is higher (Arrhenius), and the rupture life is shorter. The stress and temperature dependence of the rupture life is often represented by a time-temperature parameter (Larson-Miller, Manson-Haferd) that combines the stress, temperature and rupture time into a single correlation. The rupture data from short-duration high-stress or high-temperature tests are extrapolated to the service condition (long-duration, lower stress, lower temperature).
Relationship to the creep curve
The creep rupture life is the time coordinate of the end of the creep curve. The rupture strain (the strain at fracture) is the strain coordinate. The rupture life and the rupture strain are both material- and condition-dependent. The Monkman-Grant relationship relates the minimum creep rate to the rupture life — their product is approximately constant for a given material. This provides a cross-check between deformation and rupture data. The rupture strain is typically in the range of a few percent to tens of percent, depending on the material and the ductility. Materials with low rupture ductility (brittle creep rupture) may fail with little warning; materials with high rupture ductility show significant deformation before fracture.
Engineering significance
Creep rupture is the primary life-limiting mechanism for many high-temperature components. The design life must be less than the creep rupture life with an adequate safety margin. The safety margin is typically expressed as a factor on stress (e.g. the design stress is 2/3 of the stress that causes rupture in the design life) or a factor on life (e.g. the design life is 1/3 of the rupture life). The margin accounts for the uncertainty in the creep properties, the material scatter and the service conditions. The rupture assessment must be made at the critical location (which may not be the location of maximum elastic stress, due to stress redistribution).
Rupture life versus structural serviceability
Creep rupture is an ultimate limit state, but it is not always the first high-temperature limit state that governs a component. Long before rupture, accumulated creep strain can change clearances, alignment, seal compression, bearing load distribution or the load path through a redundant structure. Local thinning, ovalisation or distortion may also become unacceptable while the material still has substantial nominal rupture life remaining. A defensible assessment therefore separates rupture life from deformation-controlled serviceability. The analyst should identify the local quantities that matter to function, track both creep strain and damage, and check whether redistribution transfers load into neighbouring features. This distinction is particularly important in restrained assemblies, rotating hardware, joints and thin sections where small dimensional changes can create secondary loads. Rupture correlations should therefore be used alongside strain, displacement and stability checks rather than treated as a complete definition of remaining structural life.
Multiaxial stress and local damage concentration
Uniaxial rupture data are the foundation of most creep-life methods, whereas real components commonly contain multiaxial stress states at notches, weld toes, section changes and constrained interfaces. Equivalent-stress measures may reproduce the broad stress dependence of creep rate, but rupture damage can be more sensitive to hydrostatic stress, principal stress and local constraint than a simple von Mises interpretation implies. The appropriate multiaxial treatment depends on the constitutive and damage model being used and must be supported by material evidence. In practical FEA, the critical question is not simply where the initial elastic stress is highest, but where sustained high temperature, constraint and long-term redistributed stress combine to promote damage. Local peaks that rapidly relax may be less significant than moderately high stresses that persist for years. Review the time history of the stress state, not only the first increment or a single contour plot.
Creep rupture is a time-dependent failure mechanism that cannot be predicted by a static analysis. The rupture life depends on the stress, temperature, time and material. Always use a creep rupture assessment (constitutive model with damage, or a parametric rupture correlation) for high-temperature components — do not rely on a static stress limit.