Jove
Visualize
Contact Us

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Domain-Adaptive Graph Attention Semi-Supervised Network for Temperature-Resilient SHM of Composite Plates.

Sensors (Basel, Switzerland)·2025
Same author

Lamb Wave Detection for Structural Health Monitoring Using a ϕ-OTDR System.

Sensors (Basel, Switzerland)·2022
Same author

Numerical Investigation on Guided Waves Dispersion and Scattering Phenomena in Stiffened Panels.

Materials (Basel, Switzerland)·2022
See all related articles
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Oct 9, 2025

Data Acquisition Protocol for Determining Embedded Sensitivity Functions
07:46

Data Acquisition Protocol for Determining Embedded Sensitivity Functions

Published on: April 20, 2016

6.3K

Damage Detection in Flat Panels by Guided Waves Based Artificial Neural Network Trained through Finite Element

Donato Perfetto1, Alessandro De Luca1, Marco Perfetto1

  • 1Department of Engineering, University of Campania "L. Vanvitelli", Via Roma 29, 81031 Aversa, Italy.

Materials (Basel, Switzerland)
|December 24, 2021
PubMed
Summary

This study demonstrates a computer-based method to find structural damage in flat panels. By using simulated data from virtual models to train an intelligent algorithm, researchers can accurately identify the location of defects in both metal and composite materials without needing extensive physical testing.

Keywords:
Artificial Neural Network (ANN)Finite Element Analysis (FEA)Structural Health Monitoring (SHM)compositesdamage detectionguided wavesmetalsstructural health monitoringdamage localizationfinite element analysiscomposite materials

Frequently Asked Questions

More Related Videos

Design and Application of a Fault Detection Method Based on Adaptive Filters and Rotational Speed Estimation for an Electro-Hydrostatic Actuator
06:45

Design and Application of a Fault Detection Method Based on Adaptive Filters and Rotational Speed Estimation for an Electro-Hydrostatic Actuator

Published on: October 28, 2022

1.8K
Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population
09:02

Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population

Published on: January 31, 2025

735

Related Experiment Videos

Last Updated: Oct 9, 2025

Data Acquisition Protocol for Determining Embedded Sensitivity Functions
07:46

Data Acquisition Protocol for Determining Embedded Sensitivity Functions

Published on: April 20, 2016

6.3K
Design and Application of a Fault Detection Method Based on Adaptive Filters and Rotational Speed Estimation for an Electro-Hydrostatic Actuator
06:45

Design and Application of a Fault Detection Method Based on Adaptive Filters and Rotational Speed Estimation for an Electro-Hydrostatic Actuator

Published on: October 28, 2022

1.8K
Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population
09:02

Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population

Published on: January 31, 2025

735

Area of Science:

  • Structural health monitoring within civil engineering
  • Computational mechanics and Artificial Neural Networks applications

Background:

Current structural health monitoring strategies often require extensive physical testing to identify material defects. Such experimental campaigns demand significant time and financial resources for data collection. Researchers have sought computational alternatives to mitigate these logistical burdens. Artificial Neural Networks have emerged as a viable solution for automated damage identification tasks. However, training these networks typically necessitates vast datasets that are difficult to obtain through traditional laboratory methods. Finite Element analysis offers a potential pathway to generate synthetic training data efficiently. No prior work had resolved the challenge of balancing simulation accuracy with network performance across diverse material types. This gap motivated the development of a hybrid approach combining numerical modeling with machine learning architectures.

Purpose Of The Study:

The aim of this work is to propose a guided wave-based Artificial Neural Network for determining the position of structural damages. Researchers seek to address the high costs associated with traditional experimental data collection. By leveraging the Finite Element Method, the study intends to create a more efficient training pipeline for machine learning models. The authors focus on developing a robust modeling technique that accurately reflects physical structural behavior. This effort addresses the need for reliable, automated tools in structural health monitoring applications. The team investigates whether synthetic data can adequately replace physical measurements for training complex algorithms. They also explore the adaptability of the network across different material types, including aluminum and composites. Ultimately, the project seeks to establish a validated framework for precise damage localization in flat panels.

Main Methods:

Review Approach framing involves the systematic development of a numerical framework for damage detection. The researchers first construct a detailed virtual representation of the flat panels. They then validate this numerical environment by comparing outputs against established analytical benchmarks. Once verified, the team generates a comprehensive dataset representing various structural conditions. This synthetic information serves as the foundation for training the machine learning architecture. The process includes testing the network on aluminum substrates to establish baseline performance metrics. Finally, the investigators evaluate the trained system on composite materials to ensure broader applicability across different industrial configurations.

Main Results:

Key Findings From the Literature indicate that the proposed system achieves high accuracy in localizing defects across all tested scenarios. The researchers report that their numerical framework successfully identifies damage positions in both aluminum and composite panels. By utilizing synthetic data, the team demonstrates that the network effectively learns to interpret wave propagation patterns. The study confirms that the model predictions align closely with both analytical and experimental data points. This high level of precision remains consistent even when the system encounters varying damage configurations. The results show that the integration of numerical modeling significantly streamlines the development phase. The findings highlight the robustness of the architecture in diverse structural environments. These outcomes suggest that the trained algorithm reliably detects structural anomalies without requiring exhaustive physical testing campaigns.

Conclusions:

The authors demonstrate that their hybrid computational framework successfully identifies damage locations across different material types. Synthesis and implications suggest that numerical simulations effectively replace costly physical testing for training intelligent algorithms. The researchers confirm that their model maintains high precision when applied to both aluminum and composite structures. This work validates the use of synthetic data for robust structural health monitoring systems. The findings indicate that the proposed architecture performs reliably under various defect configurations. The study provides a scalable methodology for monitoring structural integrity in complex panels. These results highlight the potential for integrating advanced numerical techniques into automated diagnostic tools. The authors conclude that their approach offers a practical solution for reducing experimental overhead in engineering applications.

The researchers propose a guided wave-based Artificial Neural Network. This system identifies the specific location of structural defects by analyzing wave propagation patterns within flat panels, which are then processed by the trained algorithm to pinpoint damage coordinates.

The authors utilize Finite Element Analysis to generate synthetic training data. This computational tool allows for the simulation of wave behavior in materials, providing the necessary information to teach the algorithm without relying solely on physical experiments.

The researchers state that Finite Element model accuracy is necessary to ensure reliable network performance. They verified this by comparing simulated predictions against both analytical calculations and physical experimental data to confirm the model's validity.

The study uses aluminum plates for initial network development and training. Subsequently, the researchers verify the system's effectiveness by applying it to composite plates, demonstrating the versatility of the data-driven approach across different material compositions.

The authors measure the precision of damage localization across various defect configurations. Their findings indicate that the system maintains a high level of accuracy in every tested scenario, regardless of the specific damage geometry or plate material.

The researchers propose that their methodology significantly reduces the costs associated with laborious experimental campaigns. They imply that this computational strategy provides a viable, efficient alternative for developing structural health monitoring tools in industrial settings.