Artificial neural network approach was used to predict the thicknesses of total (FeB+Fe2B), FeB and Fe2B borides layers of AISI 1020, AISI 1060, and AISI 4140 steels. Boronizing heat treatment was conducted in a solid medium comprising of EKabor®2 powders at 840–960 ˚C at 40 ˚C intervals for 2, 4, 6, and 8 hours. Optical microscope analysis of the borided layer revealed the saw-tooth (columnar) and planar morphology. The depth of the total (FeB+Fe2B), FeB and Fe2B boride layers was accurately predicted. For total boride layers generated by the artificial neural network model, the average error varied between 0.04 and 7.64 µm. Micro hardness values increased by 423% in AISI 1020, 336% in AISI 1060, and 411% in AISI 41040 after the boronizing process.
Artificial neural network approach was used to predict the thicknesses of total (FeB+Fe2B), FeB and Fe2B borides layers of AISI 1020, AISI 1060, and AISI 4140 steels. Boronizing heat treatment was conducted in a solid medium comprising of EKabor®2 powders at 840–960 ˚C at 40 ˚C intervals for 2, 4, 6, and 8 hours. Optical microscope analysis of the borided layer revealed the saw-tooth (columnar) and planar morphology. The depth of the total (FeB+Fe2B), FeB and Fe2B boride layers was accurately predicted. For total boride layers generated by the artificial neural network model, the average error varied between 0.04 and 7.64 µm. Micro hardness values increased by 423% in AISI 1020, 336% in AISI 1060, and 411% in AISI 41040 after the boronizing process.
Birincil Dil | İngilizce |
---|---|
Konular | Yapay Zeka (Diğer), Malzeme Tasarım ve Davranışları |
Bölüm | Makaleler |
Yazarlar | |
Erken Görünüm Tarihi | 29 Mart 2024 |
Yayımlanma Tarihi | 29 Mart 2024 |
Gönderilme Tarihi | 11 Kasım 2023 |
Kabul Tarihi | 4 Mart 2024 |
Yayımlandığı Sayı | Yıl 2024 |