Multiscale-constraint based model to predict uniaxial/multiaxial creep damage and crack growth in 316-H steels.

• A new unifying analytical approach to predict creep rupture and crack growth rates by deriving a multiscale strain-based constraint criterion. • In essence, the model links the global geometric constraint with time-dependent microstructural constraint arising from creep damage at the sub grain. • By modelling the tensile and the short-term tensile properties and long-term creep strength of the material a micro constrain based argument is developed. • The model has been validated using large data sets for 316 steels at different temperatures by considering bounds in the failure strain scatter which is substantial especially at 550 °C. • It is an ideal tool for the engineer to predict component failure using simply measured data for creep strain and tensile stress. A new failure ductility/multiscale constraint strain-based model to predict creep damage, rupture and crack growth under uniaxial and multiaxial conditions is developed for 316H Type stainless steels by linking globally uniform failure strains with a multiaxial constraint factor. The model identifies a geometric constraint and a time-dependent local constraint at the sub-grain level. Uniaxial and notched 316H steel as-received and pre-compressed data at various load levels and temperatures with substantial scatter were used to derive the appropriate constitutive equations by using the proposed empirical/mechanistic approach. Constrained hydrostatic development of creep damage at the sub-grain level is assumed to directly relate to the uniform lower-bound creep steady state region of damage development measured at the global level. Uniaxial and notched bar rupture at long terms is predicted based on the initial short-term creep or a representative tensile strength and a multiaxial constraint factor. The model is consistent with the well-known NSW remaining multiaxial ductility creep crack growth model which predicts crack growth bounds over the plane strain/stress states. This model, therefore, unifies the creep process response over the whole range of uniaxial, notched and crack growth processes which is extremely consequential to simple long term failure predictions of components at elevated temperatures. [ABSTRACT FROM AUTHOR]