Abstract:Modern structures on aircraft make increasing use of large-scale composite structures. Quantitative damage monitoring for composites, including damage occurrence, number, localization, and size estimation, will help reduce maintenance costs, improve fleet management efficiency through condition-based maintenance, and potentially more rapidly enable new material systems and structural concepts by integrating health monitoring into the design itself. With the advantage of easily interpretable, intuitive, and accurate imaging result, the delay-and-sum imaging algorithm is frequently researched and applied to damage monitoring. However, when it is applied to multidamage monitoring in large-scale composites, the consumed time and computation resource are too much for pixel value calculation. Besides, due to the material anisotropy and existence of bolt holes and stiffeners in the researched complex carbon fiber composite laminate, propagation mechanism of Lamb wave is quite complicated, which makes the conventional localization result not accurate by delay-and-sum algorithm according to the point with pixel peak value. What is more, the imaging quality is deteriorated with concurrence of multiple damages, and thus, quantitative damage information cannot be extracted. Hence, the damage index merging algorithm is introduced for quick damage identification and damage merging in every subarea divided by piezoelectric sensor array. The delay-and-sum imaging algorithm is performed afterward only in subareas identified with damages, which significantly improves the efficiency of damage imaging for large-scale composites. The nonlinear normalization of pixel values compensates the deterioration of multidamage imaging quality. Then, a weighted average algorithm is introduced for more accurate localization. A further damage size level estimation is realized with probabilities by extracting the image pixel peak value. Experiments with six damages simultaneously on the complex carbon fiber laminate verify the effectiveness of the proposed quantitative multidamage monitoring algorithm with localization error below 2 cm and correct damage number and size level estimation.

Multiscale 1-DCNN for Damage Localization and Quantification Using Guided Waves with Novel Data Fusion Technique and New Self-Attention Module

Complex CFRP damage detection with dual multiscale 1DCNNs and self-selective fusion technique of guided waves

Damage Localization for Composite Structure Using Guided Wave Signals with Gramian Angular Field Image Coding and Convolutional Neural Networks

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

ConvMPAnet: A Novel End-to-End Lightweight Damage Localization Framework under Heavy Noise in Composite Structures

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

Damage Localization in Composite Structures Based on Lamb Wave and Modular Artificial Neural Network

Damage Localization with Lamb Waves Using Dense Convolutional Sparse Coding Network

Damage Quantification Method for Aircraft Structures Based on Multi-Task CNN-LSTM and Transfer Learning

A Multimodal Gated Recurrent Unit Neural Network Model for Damage Assessment in CFRP Composites Based on Lamb Waves and Minimal Sensing

Damage Quantification Method for Aircraft Structures Based on Multitask CNN-LSTM and Transfer Learning

A Guided-Wave Based Damage Diagnosing Method with Energy Spectrum and Multi-scale Network

Global-local Feature Cross-Fusion Network for Ultrasonic Guided Wave-Based Damage Localization in Composite Structures

Deep Convolutional Neural Network-Based Structural Damage Localization and Quantification Using Transmissibility Data

Damage characterization using CNN and SAE of broadband Lamb waves

Application of Artificial Neural Networks for Quantitative Damage Detection in Unidirectional Composite Structures Based on Lamb Waves

A framework of structural damage detection for civil structures using a combined multi-scale convolutional neural network and echo state network

A quantitative multidamage monitoring method for large-scale complex composite

A hybrid methodology for structural damage detection uniting FEM and 1D-CNNs: Demonstration on typical high-pile wharf

Advanced structural health monitoring in carbon fiber-reinforced plastic using real-time self-sensing data and convolutional neural network architectures

Interpretable Convolutional Sparse Coding Method of Lamb Waves for Damage Identification and Localization