Fitness-for-service (FFS) assessments are critical to the integrity management of offshore and subsea assets. Decisions regarding continued service life or the need for corrective action for damaged structures pivot on accurate FFS assessment results. While FFS assessments using failure assessment diagrams (FAD) and finite element method (FEM) have been successfully implemented on simple and regular geometries, they are not suitable for structures with complex geometries, transition of failure modes, presence of residual stress, and nonlinear fracture toughness. Extended finite element method (XFEM), a fracture mechanics-based approach enriched by extra functions around a crack, is capable of considering the above mentioned scenarios and evaluating the crack behavior. This paper demonstrates the performance of XFEM and validates the results obtained from XFEM.
First, XFEM is implemented in assessing a stationary crack on ASTM Compact Test (CT) specimen to calculate the stress intensity factor (SIF) which the obtained results deviate from the analytical solution by less than 6% for various crack length cases. Following that, a cracked plate case treated with cold expansion technique is investigated. Its remaining fatigue life is obtained by simulating fatigue crack growth, under two sets of residual stresses generated by different mandrel diameters. The results are then compared to the crack arrest hole (CAH) approach.
Through these case studies, XFEM shows adequacy for FFS applications. XFEM facilitates the modeling of the crack surface, and eliminates the need to remesh for crack growth analysis. Arbitrary structure geometries and loading combinations can be directly used in XFEM since the stress and strain responses are calculated in a conventional FEM framework. This means that the presence of local corrosion and dents, as well as transition of failure modes can be accounted for. The residual stress effect can be accurately calculated and considered for SIF calculation. Although XFEM appears to be a good solution for FFS application, adequate caution should be given to the mesh size selection and mesh orientation because they may cause slight or noticeable fluctuation of results. Therefore, a mesh sensitivity study is recommended.