Scholarlink

<p>This study investigates the indentation behavior of stitched hybrid carbon-glass quadraxial laminates with varying thickness under hemispherical, conical and flat indenters. The experimental results show peak load and energy absorption increased with laminate thickness, where thicker laminates exhibited an increase of 30–40%. A transition in damage behavior was observed for CGT-3 laminate where failure modes transitioned from matrix cracking and fiber fracture in thin laminates to delamination and shear plugging in thicker laminates. Flat indenters produced a larger damage area due to through thickness shear, whereas hemispherical and conical indenters resulted in more distributed damage zones. A machine -learning based model was developed to predict indentation response. Multilayer perceptron model (MLP) accurately predicted displacement with R<sup>2</sup> ~ 0.90, while Extra Trees Regressor model showed superior performance for force and displacement prediction where R<sup>2</sup> ~ 0.97–0.98. The results demonstrate that machine learning can be effectively used to predict indentation response and damage behavior of stitched composite laminates with good accuracy.</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

Effect of indenter geometry and machine-learning-based prediction on the quasi-static indentation response of stitched carbon-glass quadraxial laminates

  • 期刊论文

  • 发表日期：

    2026年4月24日

查看全文
  • Avinash Thirunavukkarasu,
  • Chandrasekar Muthukumar,
  • Sanjay Dakshinamoorthy

摘要

This study investigates the indentation behavior of stitched hybrid carbon-glass quadraxial laminates with varying thickness under hemispherical, conical and flat indenters. The experimental results show peak load and energy absorption increased with laminate thickness, where thicker laminates exhibited an increase of 30–40%. A transition in damage behavior was observed for CGT-3 laminate where failure modes transitioned from matrix cracking and fiber fracture in thin laminates to delamination and shear plugging in thicker laminates. Flat indenters produced a larger damage area due to through thickness shear, whereas hemispherical and conical indenters resulted in more distributed damage zones. A machine -learning based model was developed to predict indentation response. Multilayer perceptron model (MLP) accurately predicted displacement with R2 ~ 0.90, while Extra Trees Regressor model showed superior performance for force and displacement prediction where R2 ~ 0.97–0.98. The results demonstrate that machine learning can be effectively used to predict indentation response and damage behavior of stitched composite laminates with good accuracy.