Failure analysis of corroded high-strength pipeline subject to hydrogen damage based on FEM and GA-BP neural network

• Published in: International journal of hydrogen energy Vol. 47; no. 7; pp. 4741 - 4758
• Main Authors: Zhang, Han, Tian, Zhigang
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 22.01.2022
• Subjects: Artificial neural networks; Burst model; Hydrogen damage; Hydrogen pipeline; Interacting corrosions; Residual strength; Burst model; Interacting corrosions; Hydrogen pipeline; Hydrogen damage; Artificial neural networks; Residual strength
• DOI: 10.1016/j.ijhydene.2021.11.082
• ISSN: 0360-3199

The pipeline is a major approach to achieving large-scale hydrogen transportation. Hydrogen damage can deteriorate the material performance of the pipe steel, like ductility and plasticity reduction. Corrosion is dominating damage that impairs a pipeline's bearing capacity and structural reliability. However, previous research barely investigated the effect of hydrogen damage on failure behaviors, residual strength and interacting effect between adjacent corrosions of corroded high-strength pipelines transporting hydrogen. Besides, hardly any burst pressure model considers hydrogen damage. In this paper, several approaches, including the finite element method (FEM), regression analysis, the orthogonal test method, and the artificial neural network method, are applied to fill the gap. First, a series of finite element models with different geometric features and hydrogen damage is established to investigate the effects of hydrogen damage and corrosion on failure behaviors and residual strength. The results show that hydrogen damage can change the corroded pipeline's failure behaviors and reduce the residual strength. Second, based on the simulation results and regression analysis, a new burst model is developed to consider the hydrogen damage and improve the estimation accuracy. Third, based on the genetic algorithm (GA), a GA-BP neural network is established and trained for accurate and efficient residual strength estimation considering hydrogen damage. Furthermore, an orthogonal test is designed and performed to investigate the effects of critical parameters on the burst pressure of the corroded pipeline after hydrogen damage. The results indicate that hydrogen damage and corrosion length have similar contributions to the residual strength. Finally, the simulation results of pipelines with multiple corrosions show that hydrogen damage has a significant impact on the interacting effect between adjacent corrosions. The results obtained are valuable for further integrity management of steel pipelines carrying hydrogen. •Failure analysis of corroded high-strength pipeline subject to hydrogen damage.•FEM and artificial neural network are used.•The interacting effect between adjacent corrosions is considered.•A new burst model considering hydrogen damage is developed.