Development of an Advanced 3D Profile Matching Approach for Pipeline Dent FEA Assessment

Abstract Mechanical damage, particularly in the form of dents, is identified as a major threat to the structural integrity of pipelines. Finite Element Analysis (FEA) modeling has gained prominence for its reliability in evaluating the fitness-for-service (FFS) compared to the more conservative Level 2 assessments. During the dent FEA modeling process, acceptable dent geometry profile agreement between the FEA model and In-Line Inspection (ILI) data is critical to capture the appropriate stresses and strains associated with the dent formation in the FEA model. Conventional FEA modeling approaches commonly utilize rigid ring, ball, idealized rigid shell shape indenters, or a combination of multiple rigid shells as the indenters. The resulting dent profiles from the FEA model are then compared with ILI caliper data through two-dimensional (2D) profile matching. This approach is adequate for most of simple shape or symmetrical shaped dents. However, for dents with complex morphologies, the 2D profile matching method can lead to significant variations between the FEA modeled and ILI measured dent shape, leading to inaccurate stress and strain distributions in the dented region. Additionally, the conventional dent FEA modeling approach is inefficient as it requires extensive iterative adjustments to indenter shape and indentation depth. This paper presents an advanced approach to customize indenter shape for dent FEA modeling by integrating machine learning techniques, advanced indenter shape adjustment algorithms, and 3D visualization techniques. This approach ensures a satisfactory 3D dent profile matching between the FEA and ILI profiles whilst significantly reducing the time and effort required for modeling adjustments.