Investigation of the mechanical behavior of AL7075 plate supported hybrid composite plates using artificial neural networks algorithm

• Published in: International journal of polymer analysis & characterization Vol. 30; no. 5; pp. 543 - 562
• Main Authors: Tepehan, Burhan, Sülü, İsmail Yasin
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 04.07.2025
• Subjects: artificial neural networks; Hybrid composites; layered composites; materials testing; vacuum infusion
• ISSN: 1023-666X; 1563-5341
• DOI: 10.1080/1023666X.2025.2491030

The mechanical behavior of the hybrid structure formed by placing an AL7075 plate as the middle layer between different composite fibers was examined. Glass fiber and carbon fiber were preferred as fibers. Epoxy was used as matrix material. Four different cases with different fiber material layer alignments were examined. The produced plates were cut according to ASTM standards suitable for the tests to be performed, and samples were created. The samples were subjected to tensile tests, three-point bending tests, and ballistic tests. It has been determined that samples produced in sequential order with different layers reached high stress values in tensile tests and bending tests. It was observed that all alignments gave successful results in ballistic tests. In layered hybrid structures, the mechanical effects of the layer order and the type of material used on the hybrid structures have been demonstrated. The Levenberg-Marquardt algorithm with artificial neural networks was applied to investigate the appropriateness of the results. The results were presented to be appropriate in the graphs created with artificial neural networks, and it could be said that they were compatible. It can be said that more effective results are obtained in the combinations of carbon/glass/carbon and glass/carbon/glass composite fibers in layer arrangements.