Scholarlink

<p>The influence of longitudinally aligned interacting corrosion defects on the failure pressure of API 5L X120 pipe elbows under internal pressure was investigated in this work. The impact of defect geometry, specifically depth, length, and spacing, on failure pressure was measured using verified Finite Element Method (FEM) simulations and artificial neural networks (ANN). The most important geometry was defect depth; failure pressure dropped by up to 66.8% at the intrados as depth increased. High accuracy with R<sup>2</sup> values of 0.99 and mean squared error of 0.000319 was obtained by developing ANN models for defects found at the intrados, crown, and extrados. With a maximum variation of 9.23%, empirical equations obtained from the trained ANNs showed strong agreement with FEM results, hence, fit for fast and conservative failure pressure calculation in engineering applications.</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

Failure pressure prediction of an API 5L X120 pipe elbow with longitudinally aligned interacting corrosion defects subjected to internal pressure

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月1日

查看全文
  • Sangetha Devi Vijaya Kumar,
  • Suria Devi Vijaya Kumar,
  • Saravanan Karuppanan,
  • Mark Ovinis,
  • Veeradasan Perumal

摘要

The influence of longitudinally aligned interacting corrosion defects on the failure pressure of API 5L X120 pipe elbows under internal pressure was investigated in this work. The impact of defect geometry, specifically depth, length, and spacing, on failure pressure was measured using verified Finite Element Method (FEM) simulations and artificial neural networks (ANN). The most important geometry was defect depth; failure pressure dropped by up to 66.8% at the intrados as depth increased. High accuracy with R2 values of 0.99 and mean squared error of 0.000319 was obtained by developing ANN models for defects found at the intrados, crown, and extrados. With a maximum variation of 9.23%, empirical equations obtained from the trained ANNs showed strong agreement with FEM results, hence, fit for fast and conservative failure pressure calculation in engineering applications.