A coupled stress-triaxiality-dependent damage viscoplasticity model on crack initiation and propagation in high-strength rail steel

2020 
Abstract In this paper, a continuum damage enhanced formulation of effective (undamaged) material properties is proposed to predict fracture behaviour of various notched tensile specimens and a compact tension (CT) specimen for a high-strength rail steel that is commonly known to be insensitive to change in hydrostatic stress. The presented constitutive model considers a thermodynamically nonlinear, isotropic continuum damage, and stress-triaxiality-dependent effective stress using a viscoplastic regularization method to reduce strain and damage localization at the sharp notch tip. Conventional loading-unloading tensile test for damage measurement and monotonic tensile test on a notched specimen are used to calibrate the constitutive equation which consists of a set of eight model parameters. Further validation is performed on the geometry transferability of the model parameters to three different notched tensile specimens in the high stress triaxiality regime. The calibrated constitutive model is then applied to simulate crack initiation and propagation in the mode I plane-strain condition, where the computed result of load versus crack mouth opening displacement shows good agreement with those determined from CT specimens. The study concludes that the critical damage parameters for fracture are relatively constant among different notched specimens of high-strength rail steel, and thus as a proper fracture criterion for crack initiation at and crack propagation from the sharp notch tip. The study also found that the complementing mechanisms for the stress-triaxiality-dependent damage evolution and strain-hardening-induced enhancement of effective matrix material properties, respectively, result in the overall hydrostatic-stress insensitivity, and thus is often regarded as a material that belongs to the classical J2 plasticity.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    67
    References
    0
    Citations
    NaN
    KQI
    []