Application of artificial neural networks for quantitative damage detection in unidirectional composite structures based on Lamb waves

Qian, Chenggen; Ran, Yunmeng; He, Jingjing; Ren, Yi; Zhang, Weifang; Wang, Rongqiao

doi:10.1177/1687814020914732

Cited by 32 publications

(15 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent research has focused on damage quantitative estimation by using Lamb wave signals features in combination with ANN and FEM in order to ensure the accuracy of ANN (strictly related to the data used to train the network) [10,16]. A Lamb-wave-based damage evaluation method assisted by an ANN model was presented by Qian et al [19], using Damage Indexes related to changes in amplitude and phase of recorded signals. They found a good agreement between experimental and predicted results in a carbon fiberreinforced polymer composite plate.…”

Section: Introductionmentioning

confidence: 99%

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

et al. 2021

View full text Add to dashboard Cite

Artificial Neural Networks (ANNs) have rapidly emerged as a promising tool to solve damage identification and localization problem, according to a Structural Health Monitoring approach. Finite Element (FE) Analysis can be extremely helpful, especially for reducing the laborious experimental campaign costs for the ANN development and training phases. The aim of the present work is to propose a guided wave-based ANN, developed through the use of the Finite Element Method, to determine the position of damages. The paper first addresses the development and assessment of the modeling technique. The FE model accuracy was proven through the comparison of the predicted results with experimental and analytical data. Then, the ANN was developed and trained on an aluminum plate and subsequently verified in a composite plate, as well as under different damage configurations. According to the results herein proposed, the ANN allowed to detect and localize damages with a high level of accuracy in all cases of study.

show abstract

Section: Introductionmentioning

confidence: 99%

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

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Compared to unsupervised learning, supervised learning methods show higher performance and wider application prospects when some prior knowledge is available. ANNs, as one of the earliest proposed supervised deep learning methods, are employed for damage detection, impact force reconstruction and impact localization [ 15 , 24 ]. ANNs are a mathematical model which simulates a biological neural system, and hence has the ability to deal with non-linear problems.…”

Section: Machine Learning Methodologiesmentioning

confidence: 99%

“…As a result, machine learning methods, particularly neural networks, have been widely applied to extract useful information from large data sets efficiently. Some of the neural networks, like artificial neural networks (ANNs) and convolutional neural networks (CNNs), have been successfully implemented in SHM, especially in damage detection, impact localization and impact classification [ 15 , 16 , 17 , 18 , 19 , 20 ], as well as the corresponding sensor optimization [ 21 ].…”

Section: Introductionmentioning

confidence: 99%

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Seno

Khodaei

2022

Polymers

View full text Add to dashboard Cite

This paper proposes a novel method for multi-class classification and uncertainty quantification of impact events on a flat composite plate with a structural health monitoring (SHM) system by using a Bayesian neural network (BNN). Most of the existing research in passive sensing has focused on deterministic approaches for impact detection and characterization. However, there are variability in impact location, angle and energy in real operational conditions which results in uncertainty in the diagnosis. Therefore, this paper proposes a reliability-based impact characterization method based on BNN for the first time. Impact data are acquired by a passive sensing system of piezoelectric (PZT) sensors. Features extracted from the sensor signals, such as their transferred energy, frequency at maximum amplitude and time interval of the largest peak, are used to develop a BNN for impact classification (i.e., energy level). To test the robustness and reliability of the proposed model to impact variability, it is trained with perpendicular impacts and tested by variable angle impacts. The same dataset is further applied in a method called multi-artificial neural network (multi-ANN) to compare its ability in uncertainty quantification and its computational efficiency against the BNN for validation of the developed meta-model. It is demonstrated that both the BNN and multi-ANN can measure the uncertainty and confidence of the diagnosis from the prediction results. Both have very high performance in classifying impact energies when the networks are trained and tested with perpendicular impacts of different energy and location, with 94% and 98% reliable predictions for BNN and multi-ANN, respectively. However, both metamodels struggled to detect new impact scenarios (angled impacts) when the data set was not used in the development stage and only used for testing. Including additional features improved the performance of the networks in regularization; however, not to the acceptable accuracy. The BNN significantly outperforms the multi-ANN in computational time and resources. For perpendicular impacts, both methods can reach a reliable accuracy, while for angled impacts, the accuracy decreases but the uncertainty provides additional information that can be further used to improve the classification.

show abstract

“…With advances in neurosciences and high-capability computing devices, recent research is focused on application of machine learning (ML) algorithms based on Artificial Neural Networks (ANN) for guided wave damage identification, localization and qualification including an assessment on the probability of occurrence of damage in metallic and composite structural members. [14][15][16][17][18][19] Jiahui et al 20 utilized probabilistic imaging algorithm and statistical method to reduce the impact of composite anisotropy in Lamb wave-based damage localization and quantification in composite plate like structures. The algorithm was validated by experiments and results indicate an accurate prediction of the damage localization and quantification with an absolute error within 11 mm and 2.2 mm respectively for a sensor spacing of 100 mm.…”

Section: Introductionmentioning

confidence: 99%

Detection and Quantification of Asymmetrically Located Structural Damages by Mode Converted Guided Waves Using Piezo Electric Elements

Patnaik¹,

Renji²,

Gopal³

2023

The International Journal of Acoustics and Vibration

View full text Add to dashboard Cite

Though damage identification using guided waves generated using ultrasonics is well proven, its usage for structural health monitoring poses difficulties. Piezo electric actuation and sensing overcomes this difficulty to some extent. In this work, usage of such guided waves for damage identification is investigated. Piezo electric wafer transducers are used for generating and sensing the guided waves. Presence of multiple modes and comparatively higher speeds of the guided waves throw up difficulties in damage identification. It is shown here that this problem can be addressed by considering different sensor location with respect to the damage with suitable interpretation of the results. Usage of fundamental antisymmetric (Ao) mode is found to be more suitable in localizing the damage compared to the fundamental symmetric (So) mode. Asymmetrically located damage causes mode conversion. It is demonstrated in this work that the mode converted guided wave (So) could be advantageously used for identification, localization, and quantification of the damage. Damage identification and localization schemes are evolved based on the location of the sensors with respect to the damage. It is shown that the reduction in the magnitude of the mode converted wave can be utilized for assessing the depth of the damage. 3D finite element based numerical models incorporating a PZT sensor are developed and validated with experimental results in terms of the characteristics of the waves, mode conversion due to damage and influence of the defect size on the received signals which are necessary for quantification of the damage.

show abstract

Application of artificial neural networks for quantitative damage detection in unidirectional composite structures based on Lamb waves

Cited by 32 publications

References 48 publications

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

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

Structural Health Monitoring Impact Classification Method Based on Bayesian Neural Network

Detection and Quantification of Asymmetrically Located Structural Damages by Mode Converted Guided Waves Using Piezo Electric Elements

Contact Info

Product

Resources

About