Simulations of crack extensions in arc-shaped tension specimens of uncharged and hydrogen-charged 21-6-9 austenitic stainless steels using cohesive zone modeling with varying cohesive parameters

• Published in: Engineering fracture mechanics Vol. 245; no. C; p. 107603
• Main Authors: Wu, Shengjia, Pan, Jwo, Korinko, Paul S., Morgan, Michael J.
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 15.03.2021; Elsevier BV; Elsevier
• Subjects: Arc-shaped tension specimen; Austenitic stainless steels; Cohesion; Cohesive zone modeling; Finite element method; Fracture test; Hydrogen effect; Hydrogen-based energy; Mathematical models; Parameters; Plane strain; Stainless steel; Strain analysis; Trapezoidal traction-separation law; Two dimensional analysis; Two dimensional models; Varying cohesive parameters; Cohesive zone modeling; Hydrogen effect; Fracture test; Arc-shaped tension specimen; Trapezoidal traction-separation law; Varying cohesive parameters
• ISSN: 0013-7944; 1873-7315
• DOI: 10.1016/j.engfracmech.2021.107603

•Simulate crack extensions in A(T) specimens using cohesive zone modeling approach.•Calibrate fixed cohesive parameters to fit experimental load–displacement curves.•Calibrate varying cohesive parameters to fit experimental load-crack extension curves.•Computational results with varying cohesive parameters fit experimental data well.•Compare cohesive parameters for uncharged and hydrogen-charged A(T) specimens. Crack extensions in uncharged and hydrogen-charged side-grooved A(T) specimens of conventionally forged 21-6-9 austenitic stainless steels are simulated using the cohesive zone modeling approach. Two-dimensional plane strain finite element analyses with fixed cohesive parameters are first conducted to fit the experimental load-displacement curves. Similar analyses using varying cohesive parameters as functions of the crack extension are then conducted to fit the experimental load-crack extension and crack extension-displacement curves. The computational results with varying cohesive parameters can fit very well the experimental data. The computational results also indicate that the average cohesive energy for the hydrogen-charged A(T) specimen is lower than that for the uncharged A(T) specimen.