OxCMData: Automated data processing and mesh generation for contour method residual stress evaluation

The contour method is a highly robust technique for measuring two-dimensional cross-sectional residual stress maps; however, its widespread application is severely bottlenecked by the complexities of raw data analysis and finite element mesh generation. Standard workflows require tedious manual intervention and rely heavily on expensive, proprietary computer-aided design (CAD) and meshing software to handle arbitrary part boundaries and noisy metrology data. This study presents a comprehensive, fully automated computational framework designed to bridge the gap between raw surface profilometry and implicit solvers. The developed pipeline incorporates a parallelised, point-to-point Iterative Closest Point (ICP) algorithm coupled with iterative boundary erosion to successfully mitigate electrical discharge machining (EDM) artefacts and optimise spatial alignment. Complex, non-rectangular geometric boundaries are autonomously detected using an -shape formulation, eliminating the need for CAD-based perimeter extraction. The aligned data undergoes bivariate B-spline smoothing before being natively extruded into a structured, FEniCS-compatible three-dimensional volumetric mesh, complete with strict Jacobian positivity enforcement. The fidelity and robustness of this framework are validated against three diverse, publicly available third-party datasets: a benchmark specimen, a highly noisy ENPOWER edge-welded beam, and a multi-material clad pressure vessel steel. In all cases, the autonomously generated meshes and computed stress fields demonstrate excellent quantitative agreement with established literature and independent experimental data. By eliminating the reliance on commercial intermediation, this open-source-compatible pipeline significantly expands access to and accelerates advanced residual stress metrology.