Session: DA-02-02 Design & Analysis of Piping and Components - 2

Paper Number: 106201

106201 - Response of Buried Pipelines Under Permanent Ground Movements: Physics-Informed Deep Neural Network Approach 

Long-distance buried pipelines are widely used to transport oil and gas. These energy infrastructures are inexorably constructed over harsh geological fields exposing them to geohazards which can result in both transient and permanent ground motion. The permanent ground movements caused by landslides, liquefaction, ground subsidence, slope failures, fault movement, etc., can result in large deformation-induced strain in pipelines, which is a major concern in the industry. In this regard, the safety and integrity of the pipelines can be assessed by evaluating the structural response, i.e., displacement and strain fields of the buried pipelines exposed to the permanent ground displacement. The mechanical and physical behavior of the pipelines subjected to ground movement can be described mathematically using Partial Differential Equations (PDEs) with reasonable assumptions. For instance, Euler-Bernoulli beam theory with large deformation and nonlinear soil resistance springs can be adopted to model and formulate the physical behavior, resulting in sets of complicated and indeterminable nonlinear PDEs. Various methods, including the Finite Element Method (FEM) and Finite Difference Method (FDM), can be used to solve the nonlinear PDEs and thus predict the structural response.  However, the FEM usually requires complicated simulation using costly commercial software. Also, both the FE and FD methods are mesh-dependent. Recent technological advancements in Neural Network (NN) algorithms in Machine Learning (ML) have allowed Deep Learning (DL) technology to tackle equations with high nonlinearity and high indeterminacy. Therefore, it is highly beneficial to have an inexpensive, novel, easy-to-implemented, simulation-free, and meshless method to get the response. As a result, Physics-Informed Neural Networks (PINNs), deep-learning neural networks that consider the underlying law of physics in PDEs, can be a potential strategy to deal with PDEs with high complexity. To this end, the neural networks thoroughly learn based on the incorporated physical laws together with the boundary and/or initial conditions, eliminating the need for large training datasets. On this basis, this research endeavor proposes an innovative, alternative, and practical method to evaluate the response of buried pipelines subjected to permanent ground movements. Meanwhile, the obtained structural response, applicability, and accuracy of the predicted results of the proposed method are assessed by comparison with the obtained results from FEM and FDM.

Presenting Author: Pouya Taraghi University of Alberta

Presenting Author Biography: Pouya is a Ph.D. student in Structural Engineering and is working on the reliability analysis of pipelines subjected to permanent ground movement using simulation free methods. He held his Master’s degree in Structural Engineering, and Bachelor's degree in Civil Engineering with honors and distinction from Urmia University, Urmia, Iran. Pouya's research interests span across the pipelines analysis, reliability analysis, deep-learning framework, numerical modeling, stability of structures, thin-walled structures, strengthening, repair, and rehabilitation of structures, and experimental investigation. Pouya's master research was about investigating the performance of CFRP-strengthened conical shells subjected to external pressure. His research findings have been recognized internationally through top world-class scientific journal publications. Pouya has worked as an adjunct lecturer at Payame Noor University, Urmia, Iran, for two years and as a research and teaching assistant at Urmia University and University of Alberta.

Authors:

Pouya Taraghi University of Alberta
Yong Li University of Alberta
Nader Yoosef-Ghodsi Enbridge Inc.
Matt Fowler Enbridge Inc.
Muntaseer Kainat Enbridge Inc.
Samer Adeeb University of Alberta