Scholarlink

<p>In oil and gas pipelines, circumferential welds frequently fail owing to plastic damage. Despite extensive research on microstructure and macromechanical behavior of circumferential welds, quantitative analysis of plastic damage incorporating microstructural mechanisms is lacking. Hence, this study investigated the plastic damage process in circumferential welds of X80 pipeline steel and developed a damage identification method integrating microstructural evolution mechanisms with strain-field properties. The strain distribution and failure behavior of the circumferential weld were analyzed through quasi-static tensile experiments combined with digital image correlation technology, whereas microstructural changes at different elongations were recorded via microscopic observation. Microstructural features were classified using convolutional neural networks and support vector machines. In addition, a generative adversarial network-based sample augmentation method was introduced, and the correlation between microstructural geometric information and microstructural features was analyzed. Finally, principal component analysis was employed to extract global strain-field features and define damage states, with a multiple linear regression model used to clarify the relationship between microstructural features and strain-field features. The results indicate that the proposed damage identification method allows the quantitative assessment of damage severity and microstructural features, making it promising for predicting the residual life of pipeline circumferential welds and providing early failure warnings.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evolution Law of Plastic Damage in Circumferential Welded Joints of X80 Pipeline Steel

  • 期刊论文

  • 发表日期：

    2025年9月30日

查看全文
  • Liu Wenhui,
  • Wang Sen,
  • Gou Jianjun,
  • Sun Xueqi,
  • Fan Lifeng

摘要

In oil and gas pipelines, circumferential welds frequently fail owing to plastic damage. Despite extensive research on microstructure and macromechanical behavior of circumferential welds, quantitative analysis of plastic damage incorporating microstructural mechanisms is lacking. Hence, this study investigated the plastic damage process in circumferential welds of X80 pipeline steel and developed a damage identification method integrating microstructural evolution mechanisms with strain-field properties. The strain distribution and failure behavior of the circumferential weld were analyzed through quasi-static tensile experiments combined with digital image correlation technology, whereas microstructural changes at different elongations were recorded via microscopic observation. Microstructural features were classified using convolutional neural networks and support vector machines. In addition, a generative adversarial network-based sample augmentation method was introduced, and the correlation between microstructural geometric information and microstructural features was analyzed. Finally, principal component analysis was employed to extract global strain-field features and define damage states, with a multiple linear regression model used to clarify the relationship between microstructural features and strain-field features. The results indicate that the proposed damage identification method allows the quantitative assessment of damage severity and microstructural features, making it promising for predicting the residual life of pipeline circumferential welds and providing early failure warnings.