A Hybrid FE-ML Approach for Critical Buckling Moment Prediction in Dented Pipelines Under Complex Loadings

Area of Science:

• Mechanical Engineering
• Structural Integrity
• Computational Mechanics

Background:

• Dents in pipelines are geometric defects that compromise structural integrity.
• Dented sections are prone to local buckling, posing significant risks to pipeline reliability.
• Accurate prediction of critical buckling moment (CBM) is crucial for ensuring pipeline safety.

Purpose of the Study:

• To develop a novel finite element (FE) machine learning (ML)-based approach for analyzing dented pipelines.
• To predict the critical buckling moment (CBM) of dented pipelines under combined internal pressure and bending moment (BM) loading.
• To quantify parametric effects on CBM and enhance buckling resistance evaluation.

Main Methods:

• A hybrid machine learning framework integrating Extreme Learning Machine (ELM) with Bald Eagle Search (BES), Lévy flight, and Simulated Annealing (SA).
• Development of a comprehensive dataset by quantifying parametric effects on CBM.
• Utilizing finite element analysis in conjunction with machine learning for predictive modeling.

Main Results:

• Achieved highly accurate predictions of the critical buckling moment (CBM).
• Demonstrated the effectiveness of the hybrid ELM-BES-Lévy-SA algorithm in CBM prediction.
• Quantified the influence of various parameters on the buckling behavior of dented pipelines.

Conclusions:

• The proposed FE-ML approach provides a reliable method for evaluating the buckling resistance of dented pipelines.
• The study offers valuable insights into the complex loading conditions affecting pipeline integrity.
• This research enhances the understanding and prediction capabilities for ensuring the safety and reliability of pipelines with dents.