A Crack Propagation Method for Pipelines with Interacting Corrosion and Crack Defects

Area of Science:

• Materials Science
• Mechanical Engineering
• Civil Engineering

Background:

• Pipeline integrity is crucial, yet often compromised by co-existing corrosion and crack defects.
• The interaction between these defects can significantly influence fatigue crack propagation, posing a risk to structural safety.

Purpose of the Study:

• To develop a novel crack propagation method for pipelines with interacting corrosion and crack defects.
• To introduce and predict the Stress Intensity Factor (SIF) interaction impact ratio using advanced machine learning techniques.

Main Methods:

• Finite element models were constructed to determine SIFs for fatigue cracks.
• Two extreme gradient boosting (XGBoost) models were developed to predict the SIF interaction impact ratio.
• The models utilized crack size, corrosion size, and axial distance as input parameters.

Main Results:

• The proposed XGBoost approaches accurately predicted the SIF interaction impact ratio using synthetic data from finite element modeling.
• Crack propagation models were successfully established by integrating the XGBoost predictions with Paris' law and a corrosion growth model.
• Sensitivity analyses confirmed the influence of initial corrosion depth and axial distance on crack growth.

Conclusions:

• The developed method effectively links crack propagation to corrosion and crack sizes, and their spatial interaction.
• This approach offers a valuable tool for enhancing pipeline integrity management by accounting for defect interactions.
• The study provides insights into the complex interplay between corrosion and crack defects on fatigue crack growth in pipelines.