Abstract:This paper reports on a novel metamodel for impact detection, localization and characterization of complex composite structures based on Convolutional Neural Networks (CNN) and passive sensing. Methods to generate appropriate input datasets and network architectures for impact localization and characterization were proposed, investigated and optimized. The ultrasonic waves generated by external impact events and recorded by piezoelectric sensors are transferred to 2D images which are used for impact detection and characterization. The accuracy of the detection was tested on a composite fuselage panel which was shown to be over 94%. In addition, the scalability of this metamodelling technique has been investigated by training the CNN metamodels with the data from part of the stiffened panel and testing the performance on other sections with similar geometry. Impacts were detected with an accuracy of over 95%. Impact energy levels were also successfully categorized while trained at coupon level and applied to sub-components with greater complexity. These results validated the applicability of the proposed CNN-based metamodel to real-life application such as composite aircraft parts.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to conduct impact detection, location, and feature identification in complex composite material structures. Specifically, the authors propose a new meta - model based on convolutional neural networks (CNN) and passive sensing techniques, aiming to solve the following problems: 1. **Impact Detection and Location**: Traditional non - destructive inspection (NDI) techniques have limitations in detecting minor damage (such as barely visible impact damage, BVID) in composite material structures, especially when continuously monitoring the structure. The paper proposes using CNN to process ultrasonic signals recorded by piezoelectric sensors and convert them into 2D images to achieve high - precision impact detection and location. 2. **Impact Feature Identification**: In addition to detection and location, this method can also identify the energy level of the impact, thereby determining whether the impact has reached the level requiring an alarm or has already caused damage. This helps to assess the health status of the structure and provides maintenance engineers with necessary action recommendations and information on the remaining service life of the structure. 3. **Model Scalability**: The research also explores the training and testing performance of this meta - model on parts with different geometries, verifying its applicability in practical applications, such as impact detection of composite aircraft components. Experimental results show that even a model trained on a small - scale structure can achieve a detection accuracy of over 95% on more complex structural components. In conclusion, by introducing CNN and passive sensing techniques, this paper aims to improve the accuracy and efficiency of impact detection in composite material structures while ensuring the scalability and practicality of the method. This is of great significance for ensuring the safety and reliability of next - generation lightweight composite aircraft.

A Convolutional Neural Network for Impact Detection and Characterization of Complex Composite Structures

Impact Force Localization and Reconstruction Via Gated Temporal Convolutional Network

Impact-Induced Damage Recognition of Aluminium Alloy Stiffened Plate Structure Based on Convolutional Neural Network

Impact localization in composite structures with Deep Neural Networks

Detection of impact on aircraft composite structure using machine learning techniques

Impact monitoring of large size complex metal structures based on sparse sensor array and transfer learning

A Hybrid FCN-BiGRU with Transfer Learning for Low-Velocity Impact Identification on Aircraft Structure

Multi-frequency probabilistic imaging fusion for impact localization on aircraft composite structures

A unified CNN approach for guided wave-based damage detection, damage size estimation and reliability assessment demonstrated on a complex composite structure

Automatic Air-Coupled Ultrasound Detection of Impact Damages in Fiber-Reinforced Composites Based on One-Dimension Deep Learning Models

Automated Debonding Characterization in Reinforced Structures Based on Finite Element Analysis and Convolutional Neural Networks

Impact identification of composite cylinder based on improved deep metric learning model and weighted fusion Tikhonov regularized total least squares

Using Acoustic Emission Technique for Structural Health Monitoring of Laminate Composite: A Novel CNN-LSTM Framework

Impact Energy Assessment of Sandwich Composites Using an Ensemble Approach Boosted by Deep Learning and Electromechanical Impedance

Structural Damage Identification Based on Convolutional Neural Networks and Improved Hunter–Prey Optimization Algorithm

Structure health monitoring of a composite wing based on flight load and strain data using deep learning method

Parallel Convolutional Neural Network Toward High Efficiency and Robust Structural Damage Identification

Study on Prediction of Compression Performance of Composite Laminates After Impact Based on Convolutional Neural Networks

Improved Damage Localization and Quantification of CFRP Using Lamb Waves and Convolution Neural Network

IMPACT MONITORING OF LARGE AND COMPLEX STRUCTURES BASED ON TRANSFER LEARNING