Profile Comparison Validation Techniques for Level 3 FEA Dent Modeling in 2 and 3 Dimensions

Abstract In-line inspection (ILI) tools contain a variety of sensors to analyze pipelines for detrimental features such as internal and external corrosion, geometric deformation, laminations, cracks, and other defects related to pipeline integrity. Level 3 assessments use finite element analysis (FEA) to replicate the indentation process to further assess the stresses and strains associated with the dent. Simulated results are extracted and analyzed and typically validated through the use of 2D profile comparisons that run through the apex of the dent between the deformed FEA model results and the ILI caliper data. These profile comparisons overlay the axial and circumferential profiles of the modeled dents with the same profiles extracted from the caliper data that is collected during ILI. Although these profiles give engineers an idea of how accurately dent morphology was captured in the FEA model, the validation approach is inherently flawed. The caliper sensors are often limited in resolution through tool error tolerances, caliper arm lift-off, and by the number of caliper arms (which can be as few as 12 spanning circumferentially). This can result in physically unrealistic 2D profiles that would be difficult to validate with a 3D model without implementing copious numbers of simulations. Additionally, as dents are 3D deformations, debilitating stresses and strains can exist outside of the peak (such as in shoulders or between adjacent dents). Therefore, relying solely on 2D validation for a 3D model might result in missing aspects of the deformation that could affect the engineer’s final assessment. This paper introduces a novel 3D profile comparison methodology incorporating advanced computational geometric techniques to measure the 3D accuracy of the modelled dent to the caliper data. The process involves implementing a measurement of the distance between a point and the nearest triangle on a representative surface mesh constructed from the ILI data. A first case study involves measuring the 3D profile between two different indentation processes to show the difference between 2D and 3D validation. The second case study then utilizes a 3D point cloud rendered from laser scanning software to a modeled indentation of the same dent. It therefore allows for a ground truth comparison between the FEA model and the actual dent morphology to determine how accurate the FEA approach is in simulating a dent.