A inserção de tecnicas de inteligencia artificial na modelagem matematica do lingotamento continuo de aços

AUTOR(ES)
DATA DE PUBLICAÇÃO

2001

RESUMO

The control of quality of products produced by continuous casting can not be achieved without a knowledge base of the process, which incorporates parameters and variables influence such as: equipment, cast metal, behavior of each component of the cooling system and operational conditions. This work presents the development of a computational algorithm (software) applied to maximize quality in the continuous casting of steel. A mathematical model of solidification connected to a genetic algorithm of search explores space parameters settings based on metallurgical criteria. Initially, the results of simulations performed with the mathematical model were validated against experimental and literature results showing a good agreement in a1l cases examined. The interaction between the mathematical model and the search method permits the selection ofthe best cooling efficiency, and the final results are presented in a report containing the ingot surface temperature and the metallurgical length. A mathematical model of solidification for the Twin-Roll strip casting was developed. The both models use a finite difference numerical scheme and permit the determination of the thermal field of the metal. Results simulated by the model were compared to data from literature models presenting a very good agreement. This work it is part of a joint project with the Institute de Pesquisas Tecnológica (IPTSP), where several studies and analysis were performed by using the Twin-Roll strip casting pilot equipment and laboratory facilities at IPT

ASSUNTO(S)

fundição continua inteligencia artificial modelos matematicos software - desenvolvimento simulação (computadores) aço - fundição algoritmos geneticos

ACESSO AO ARTIGO

Documentos Relacionados