Corroded pipeline failure analysis using artificial neural network scheme. (October 2017)

• Record Type: Journal Article
• Authors: Xu, Wen-Zheng; Li, Chun Bao; Choung, Joonmo; Lee, Jae-Myung
• Is Part Of: Advances in engineering software. Volume 112(2017)
• Page Start: 255
• Page End: 266
• Publication Date: 2017-10
• Subjects: Artificial neural networks -- Interacting corrosion defects -- API X80 pipelines -- Burst pressure; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2017.05.006
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: The failure behavior of pipelines with interacting corrosion defects was studied using FE method. A series of models were created for the sensitive study of the various parameters. A solution was proposed to predict burst pressure using an artificial neural network (ANN). The solution was validated by comparing with experimental results and existing codes. Abstract: Corrosion defects occur very often on the internal and external surfaces of pipelines, which may result in a serious threat to the integrity of the pipelines. Numerous studies investigated failure behavior of corroded pipelines with single corrosion defects. However, few studies focus on interacting corrosion defects. Interacting defects are defined as defects with certain proximity that interact to reduce the overall strength of a pipeline. In the present study, the failure behavior of pipelines with interacting corrosion defects was studied using a finite element method, and then a solution was proposed to predict burst pressure using an artificial neural network. The solution was validated by experimental results in previous studies and compared with other existing assessment solutions to prove its applicability and efficiency.