Neural-fuzzy optimization of thick composites curing process
Само за регистроване кориснике
2019
Аутори
Aleksendrić, DraganBellini, Costanzo
Carlone, Pierpaolo
Ćirović, Velimir
Rubino, Felice
Sorrentino, Luca
Чланак у часопису (Објављена верзија)
Метаподаци
Приказ свих података о документуАпстракт
This article addresses the optimization of curing process for thick composite laminates. The proposed methodology aims at the evaluation of the thermal cycle promoting a desired evolution of the degree of cure inside the material. At the same time, temperature overshooting as well as excessive temperature and cure degree gradient through the thickness of the material are prevented. The developed approach is based on the integrated application of artificial neural networks and a fuzzy logic controller. The neural networks promptly predict the behavior of composite material during curing process, while the fuzzy logic controller continuously and opportunely adjusts the proper variations on the imposed thermal cycle. The results highlighted the efficiency of the method in comparison with the cure profiles dictated by the material suppliers. For thick laminates, a reduction of 35% of cure time and improvements of approximately 10% of temperature overshooting was obtained compared to conven...tional curing cycles. The method was validated by experimental tests.
Кључне речи:
thick / process / optimization / neural / networks / logic / fuzzy / Curing / composites / artificialИзвор:
Materials and Manufacturing Processes, 2019, 34, 3, 262-273Издавач:
- Taylor & Francis Inc, Philadelphia
DOI: 10.1080/10426914.2018.1512116
ISSN: 1042-6914
WoS: 000456880400003
Scopus: 2-s2.0-85053256133
Колекције
Институција/група
Mašinski fakultetTY - JOUR AU - Aleksendrić, Dragan AU - Bellini, Costanzo AU - Carlone, Pierpaolo AU - Ćirović, Velimir AU - Rubino, Felice AU - Sorrentino, Luca PY - 2019 UR - https://machinery.mas.bg.ac.rs/handle/123456789/3032 AB - This article addresses the optimization of curing process for thick composite laminates. The proposed methodology aims at the evaluation of the thermal cycle promoting a desired evolution of the degree of cure inside the material. At the same time, temperature overshooting as well as excessive temperature and cure degree gradient through the thickness of the material are prevented. The developed approach is based on the integrated application of artificial neural networks and a fuzzy logic controller. The neural networks promptly predict the behavior of composite material during curing process, while the fuzzy logic controller continuously and opportunely adjusts the proper variations on the imposed thermal cycle. The results highlighted the efficiency of the method in comparison with the cure profiles dictated by the material suppliers. For thick laminates, a reduction of 35% of cure time and improvements of approximately 10% of temperature overshooting was obtained compared to conventional curing cycles. The method was validated by experimental tests. PB - Taylor & Francis Inc, Philadelphia T2 - Materials and Manufacturing Processes T1 - Neural-fuzzy optimization of thick composites curing process EP - 273 IS - 3 SP - 262 VL - 34 DO - 10.1080/10426914.2018.1512116 ER -
@article{ author = "Aleksendrić, Dragan and Bellini, Costanzo and Carlone, Pierpaolo and Ćirović, Velimir and Rubino, Felice and Sorrentino, Luca", year = "2019", abstract = "This article addresses the optimization of curing process for thick composite laminates. The proposed methodology aims at the evaluation of the thermal cycle promoting a desired evolution of the degree of cure inside the material. At the same time, temperature overshooting as well as excessive temperature and cure degree gradient through the thickness of the material are prevented. The developed approach is based on the integrated application of artificial neural networks and a fuzzy logic controller. The neural networks promptly predict the behavior of composite material during curing process, while the fuzzy logic controller continuously and opportunely adjusts the proper variations on the imposed thermal cycle. The results highlighted the efficiency of the method in comparison with the cure profiles dictated by the material suppliers. For thick laminates, a reduction of 35% of cure time and improvements of approximately 10% of temperature overshooting was obtained compared to conventional curing cycles. The method was validated by experimental tests.", publisher = "Taylor & Francis Inc, Philadelphia", journal = "Materials and Manufacturing Processes", title = "Neural-fuzzy optimization of thick composites curing process", pages = "273-262", number = "3", volume = "34", doi = "10.1080/10426914.2018.1512116" }
Aleksendrić, D., Bellini, C., Carlone, P., Ćirović, V., Rubino, F.,& Sorrentino, L.. (2019). Neural-fuzzy optimization of thick composites curing process. in Materials and Manufacturing Processes Taylor & Francis Inc, Philadelphia., 34(3), 262-273. https://doi.org/10.1080/10426914.2018.1512116
Aleksendrić D, Bellini C, Carlone P, Ćirović V, Rubino F, Sorrentino L. Neural-fuzzy optimization of thick composites curing process. in Materials and Manufacturing Processes. 2019;34(3):262-273. doi:10.1080/10426914.2018.1512116 .
Aleksendrić, Dragan, Bellini, Costanzo, Carlone, Pierpaolo, Ćirović, Velimir, Rubino, Felice, Sorrentino, Luca, "Neural-fuzzy optimization of thick composites curing process" in Materials and Manufacturing Processes, 34, no. 3 (2019):262-273, https://doi.org/10.1080/10426914.2018.1512116 . .