Full chain modeling from multi-source geometric errors to pipeline end face topography for subsea pipeline cutting machine

When subsea oil and gas pipelines malfunction, specialized cutting equipment is required to sever them, and the cutting accuracy of such equipment directly affects subsequent maintenance processes. Focusing on the self-developed Subsea Pipeline Cutting Machine (SPCM), this study proposes a full-chain modeling framework of “multi-source geometric errors-tool pose errors-geometric topography of machined pipeline surfaces.” Firstly, based on the multibody system (MBS) theory, a geometric error model for the SPCM main body is established, which maps the position-independent/dependent geometric Errors (PIGE/PDGE) of each component to the tool pose errors. Secondly, a mathematical model of the actual tool cutting trajectory is constructed, which incorporates both tool pose errors and tool deformation errors. Furthermore, the Z-map method is improved to propose a simulation algorithm for machined surface topography suitable for pipeline cutting. Thirdly, an improved Sobol method is employed for global sensitivity analysis to accurately identify the key error terms affecting surface topography, and the Morris method is used to verify its effectiveness. Finally, the results of pipeline cutting experiments show that under different feeds per tooth, the maximum deviation between the theoretically simulated surface roughness and the experimental results is 8.87%, which verifies the correctness of the proposed full-chain model.