A dual-scale stochastic analysis framework for creep failure considering microstructural randomness

Creep failure under high temperatures is a complex multiscale and multi-mechanism issue involving inherent microstructural randomness. To investigate the effect of microstructures on the uniaxial/multiaxial creep failure, a dual-scale stochastic analysis framework is established to introduce the grain boundary (GB) characteristics into the macroscopic analysis. The GB degeneration-dominated creep failure of nickel-base superalloy Inconel 617 under long-term creep is considered in this study. Firstly, the damage mechanisms of GBs are investigated based on the crystal plasticity finite element (CPFE) method and cohesive zone model (CZM). Subsequently, based on the obtained GB damage evolution, a novel Monte Carlo (MC) approach is proposed to quantify the randomness of macroscopic creep behavior based on the statistical feature of GB orientation and area distribution. Finally, a dual-scale stochastic multiaxial creep damage model is established to incorporate the influence of the random GB orientation and area distribution. With the numerical application of the proposed creep damage model, the random initiation and growth of creep cracks in the uniaxial tensile specimen and the pressurized tube are captured and analyzed. The proposed stochastic framework effectively considers the inherent randomness introduced by GB characteristics and efficiently realizes full-field multiscale calculations. It also shows its potential applications in safety evaluation and life prediction of creep components and structures under high temperatures.