Evaluation of a Nozzle Crack-Like Flaw Using Computational Fracture Mechanics With the API 579-1/ASME FFS-1 Annex 9H Procedure

Abstract Annex 9H in API 579-1/ASME FFS-1 Fitness-For-Service (API 579-1) contains a procedure to reduce conservatism in the calculation of the toughness ratio, Kr, in a Fitness-For-Service (FFS) Assessment of a crack-like flaw. The Annex 9H procedure incorporates constraint effects (based on T-stress, a measure of crack front constraint) to improve the fracture toughness, Kmat. Applying Annex 9H reduces the estimated Kr value in the Failure Assessment Diagram (FAD) used to determine if the crack-like flaw is acceptable or unacceptable. However, the procedure in this annex is limited in its application. Annex 9H can only be applied to surface-breaking semi-elliptical, crack-like flaws in flat plate and cylindrical shell geometries when the stress distribution across the flaw plane can be accurately represented by linearization. WRC Bulletin 590 provides a method that expands on the Annex 9H procedure for non-linear stress distributions yet is still limited to these geometries. PVP2024-124960 by Blanks proposed a procedure enabling Annex 9H to be implemented in an FFS assessment using computational fracture mechanics three-dimensional (3D) crack meshes with Finite Element Analysis (FEA). Using computational fracture mechanics enables the procedure to be applied to the general case of a surface-breaking, crack-like flaw in any component under any stress distribution. This paper describes an assessment of a crack-like flaw in a nozzle to elliptical head weld of a pressure vessel. The assessment was initially carried out using the API 579-1 Part 9 Level 2 Assessment procedures, with stresses from an FEA of the vessel in the un-cracked configuration. The Annex 9H/WRC Bulletin 590 procedures were used to incorporate constraint effects and improve the calculation of the fracture toughness. However, the necessary geometry idealization to evaluate the crack driving forces (e.g., stress intensity factors) and T-stress solutions resulted in an overly conservative result. The crack was subsequently incorporated into the stress analysis with a 3D crack mesh, enabling direct calculation of the crack driving forces and T-stresses. The PVP2024-124960 procedure was employed in the assessment to incorporate constraint effects, and the resulting calculation of the toughness ratio improved when compared to the initial assessment. The reduction in toughness ratio was a function of the improvement (i.e., reduction) in the calculation of the stress intensity factors and an increase in fracture toughness from improvement in the calculation of T-stresses. Comparisons are made between the results of the assessment following the typical Part 9 procedures (i.e., stresses analysis results based on an un-cracked configuration with geometry idealizations), with and without Annex 9H, to the results of the assessment using FEA with a 3D crack mesh, with and without Annex 9H. The comparisons highlight the benefits of both the Annex 9H procedures and the use of computational fracture mechanics when assessing a crack-like flaw in a complex geometry.