Pregled bibliografske jedinice broj: 268362
Prediction of Phase Transformation Using Neural Networks
Prediction of Phase Transformation Using Neural Networks // Proceedings of 2nd International Conference Heat Treatment and Surface Engineering in Automotive Applications
Riva del Garda, Italija, 2005. (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 268362 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Prediction of Phase Transformation Using Neural Networks
Autori
Smoljan, Božo ; Smokvina Hanza, Sunčana ; Filetin, Tomislav
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of 2nd International Conference Heat Treatment and Surface Engineering in Automotive Applications
/ - , 2005
Skup
2nd International Conference Heat Treatment and Surface Engineering in Automotive Applications
Mjesto i datum
Riva del Garda, Italija, 20.06.2005. - 22.06.2005
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
Steel; quenching; phase transformation; microstructure composition; heat treatment; neural network
Sažetak
The application of a static multi layer feed forward neural network with learning rule based on error backpropagation algorithm in prediction of microstructure composition of quenched steel has been investigated. To accelerate the convergence of the learning algorithm, the momentum method has been applied. Prediction of quenched steel microstructure composition has been estimated based on elemental composition, total hardness of quenched steel, austenitizing temperature, austenitizing time and time of cooling from 800 °C to 500 °C. The differences between predicted and measured microstructure compositions for testing steel types can be neglected. Established method is suitable for microstructure prediction of low-alloyed steels.
Izvorni jezik
Engleski
Znanstvena područja
Strojarstvo
