Prediction of Buckling Behaviour of Composite Plate Element Using Artificial Neural Networks

Prediction of Buckling Behaviour of Composite Plate Element Using Artificial Neural Networks

2024.01.15 | Katarzyna Falkowicz, Monika Kulisz
This article explores the use of artificial neural networks (ANNs) to predict the buckling behavior of composite plate elements with cut-outs, which can function as spring elements. The study combines numerical analysis with ANNs to evaluate the stability of these plates under compression. The ANNs models were trained and tested using a large dataset, and their accuracy was evaluated using various statistical measures. The results show that the developed ANNs models are highly accurate in predicting the critical force and buckling form of thin-walled plates with different cut-out and fiber angel configurations. The integration of numerical analyses with ANNs provides a practical and efficient solution for designing and monitoring composite structures, offering significant time and cost savings in research and production. The study focuses on a rectangular carbon-epoxy composite plate with adjustable geometric dimensions and demonstrates the effectiveness of ANNs in predicting buckling forms and critical forces. The models achieved high accuracy, with correlation coefficients (R) exceeding 0.999 and mean squared errors (MSE) and root mean square errors (RMSE) indicating low error metrics. The classification model also achieved 100% accuracy, as evidenced by the ROC curve and confusion matrix. The findings confirm the potential of ANNs in predicting the behavior of composite plates, which can be valuable for optimizing designs and structural monitoring in various engineering applications.This article explores the use of artificial neural networks (ANNs) to predict the buckling behavior of composite plate elements with cut-outs, which can function as spring elements. The study combines numerical analysis with ANNs to evaluate the stability of these plates under compression. The ANNs models were trained and tested using a large dataset, and their accuracy was evaluated using various statistical measures. The results show that the developed ANNs models are highly accurate in predicting the critical force and buckling form of thin-walled plates with different cut-out and fiber angel configurations. The integration of numerical analyses with ANNs provides a practical and efficient solution for designing and monitoring composite structures, offering significant time and cost savings in research and production. The study focuses on a rectangular carbon-epoxy composite plate with adjustable geometric dimensions and demonstrates the effectiveness of ANNs in predicting buckling forms and critical forces. The models achieved high accuracy, with correlation coefficients (R) exceeding 0.999 and mean squared errors (MSE) and root mean square errors (RMSE) indicating low error metrics. The classification model also achieved 100% accuracy, as evidenced by the ROC curve and confusion matrix. The findings confirm the potential of ANNs in predicting the behavior of composite plates, which can be valuable for optimizing designs and structural monitoring in various engineering applications.
Reach us at info@study.space
[slides and audio] Prediction of Buckling Behaviour of Composite Plate Element Using Artificial Neural Networks