Scholarlink

<p>This study investigates how manufacturing defects transform the statistical distribution of failure in carbon fiber-reinforced polymer composites under tension and compression loading. The analysis of tension and compression specimens reveals that defect-free composites exhibit relatively narrow, unimodal strength distributions. In contrast, specimens with porosity or fiber waviness develop more complex multimodal probability densities with fat-tailed distributions and substantially higher variability. Applying Jensen’s inequality demonstrates that this increased variability can be assessed to identify higher risk profiles. These findings indicate that defects in composites don’t simply reduce mean strength values but alter the statistical nature of composite failure, transforming thin-tailed, unimodal, well-behaved distributions into multimodal, fat-tailed ones. Such transformation necessitates more sophisticated probabilistic approaches for reliable design and strength prediction in safety-critical applications where understanding tail risk becomes crucial to proper risk management.</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

Fat-tailed failure strength distributions and manufacturing defects in advanced composites

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月17日

查看全文
  • Rani Elhajjar

摘要

This study investigates how manufacturing defects transform the statistical distribution of failure in carbon fiber-reinforced polymer composites under tension and compression loading. The analysis of tension and compression specimens reveals that defect-free composites exhibit relatively narrow, unimodal strength distributions. In contrast, specimens with porosity or fiber waviness develop more complex multimodal probability densities with fat-tailed distributions and substantially higher variability. Applying Jensen’s inequality demonstrates that this increased variability can be assessed to identify higher risk profiles. These findings indicate that defects in composites don’t simply reduce mean strength values but alter the statistical nature of composite failure, transforming thin-tailed, unimodal, well-behaved distributions into multimodal, fat-tailed ones. Such transformation necessitates more sophisticated probabilistic approaches for reliable design and strength prediction in safety-critical applications where understanding tail risk becomes crucial to proper risk management.