The accurate prediction of failure pressures in pipelines containing multiple defects is important for assessing the integrity and reliability of corroded pipelines. First, the effect of the remaining defects on the failure pressure of the pipeline containing triple defects was investigated by developing the finite element (FE) model of pipelines containing triple defects. Then, to integrate machine learning method and finite element models to establish a prediction model for the interaction coefficients corresponding to combined corrosion defects. Subsequently, a framework for predicting the burst pressure of pipelines appliable to the scenario of group corrosion defects was proposed by integrating the interacting coefficient and the failure pressure of single defect. Finally, the accuracy of the model was demonstrated by the evaluation indicators and bursting experimental data of corroded pipeline containing cluster defects. The framework is expected to provide a maintenance foundation for the integrity assessment of pipelines with clustered corrosion defects.
Read full abstract