Scholarlink

Purpose <p>This study develops a reusable non-bonded piezoceramic sensor (NBPS) system for split-tensile structural health monitoring of steel fibre concrete structures. The research addresses critical limitations in conventional destructive testing methodologies by integrating electromechanical impedance (EMI) signatures with deep learning regression for non-destructive tensile performance evaluation.</p> Methods <p>The enhanced monitoring system evaluated both healthy and damaged steel fibre concrete cylinders at six volumetric dosages (0.25–1.5%) and two aspect ratios (55, 65) after 28 days of curing. EMI signatures, including resonance frequency, conductance peak, and stiffness-loss index (Δs), were compiled into conventional tensile and processed using RMSD metrics. Equivalent structural parameters were computed via a calibrated Hixon model, and deep learning networks were trained on raw EMI data for tensile strength prediction.</p> Results <p>The maximum tensile strength of 6.68&#xa0;MPa was achieved at 1.50% fibre content and an aspect ratio of 65, representing a 65.3% enhancement over plain M30 concrete. The calibrated model confirmed predominantly stiffness-controlled damage mechanisms through equivalent spring-mass-damper system analysis. LSTM achieved exceptional prediction accuracy with R² = 0.9996 and MAE less than 0.02, significantly outperforming CNN and Bi-LSTM architectures in processing raw EMI signatures.</p> Conclusion <p>The NBPS-EMI-DL system provides fast, identical, repeatable, and non-destructive tensile diagnostics for steel fibre concrete structures. The validated correlation between EMI signatures and structural damage progression enables automated quality control and establishes pathways for closed-loop structural health monitoring of fibre-reinforced concrete infrastructure.</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

Electromechanical Impedance−Based Tensile Damage Assessment of Steel Fibre Concrete Using Reusable Piezoceramic Sensors: An Experimental and Deep Learning Approach

  • 期刊论文

  • 发表日期：

    2026年3月14日

查看全文
  • Wesam Al Agha,
  • Shilpa Pal,
  • Nirendra Dev

摘要

Purpose

This study develops a reusable non-bonded piezoceramic sensor (NBPS) system for split-tensile structural health monitoring of steel fibre concrete structures. The research addresses critical limitations in conventional destructive testing methodologies by integrating electromechanical impedance (EMI) signatures with deep learning regression for non-destructive tensile performance evaluation.

Methods

The enhanced monitoring system evaluated both healthy and damaged steel fibre concrete cylinders at six volumetric dosages (0.25–1.5%) and two aspect ratios (55, 65) after 28 days of curing. EMI signatures, including resonance frequency, conductance peak, and stiffness-loss index (Δs), were compiled into conventional tensile and processed using RMSD metrics. Equivalent structural parameters were computed via a calibrated Hixon model, and deep learning networks were trained on raw EMI data for tensile strength prediction.

Results

The maximum tensile strength of 6.68 MPa was achieved at 1.50% fibre content and an aspect ratio of 65, representing a 65.3% enhancement over plain M30 concrete. The calibrated model confirmed predominantly stiffness-controlled damage mechanisms through equivalent spring-mass-damper system analysis. LSTM achieved exceptional prediction accuracy with R² = 0.9996 and MAE less than 0.02, significantly outperforming CNN and Bi-LSTM architectures in processing raw EMI signatures.

Conclusion

The NBPS-EMI-DL system provides fast, identical, repeatable, and non-destructive tensile diagnostics for steel fibre concrete structures. The validated correlation between EMI signatures and structural damage progression enables automated quality control and establishes pathways for closed-loop structural health monitoring of fibre-reinforced concrete infrastructure.