Seleção de sistemas de isolamento térmico multi-componentes para fornos de altas temperaturas baseada no desempenho térmico

• Ano de defesa: 2019
• Autor(a) principal: Santos, Denis Prado
• Orientador(a): Pandolfelli, Victor Carlos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência e Engenharia de Materiais - PPGCEM
• Palavras-chave em Português: Materiais cerâmicos refratários; Isolantes térmicos; Forno elétrico resistivo; Método dos elementos finitos; Algoritmos genéticos; Inteligência computacional; Seleção de materiais
• Palavras-chave em Inglês: Refractory ceramic materials; Thermal insulators; Electric resistance furnace; Finite element method; Genetic algorithms; Computational inteligence; Materials selection
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE MATERIAIS E METALURGICA::MATERIAIS NAO METALICOS
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/16562

Controlled heating through furnace development has allowed the humankind to mold the matter to obtain various products for several different purposes. However, some of them require high-temperature processes so that the necessary physical and chemical reactions can occur, leading to local temperature increase and high energy consumption. Refractory ceramic materials are used for lining the walls of furnaces, in order to reduce the heat dissipation from the heating chamber, and thus minimizing the required thermal energy for keeping its high temperatures. Nevertheless, there is a huge number of commercial insulating materials available, with several properties ranges, that could be used for furnace design, making the selection of the optimal combination of materials a difficult task. This work explores the possibility of using computational tools for the modeling and optimization, providing data-driven recommendations for furnace designers to choose suitable insulating systems. Thus, an evolutionary screening procedure (ESP), based on finite element method and stochastic optimization by multi-objective genetic algorithms, was developed and evaluated on a case-study related to an electric resistance furnace operating up to 1600$^o$C. The ESP has shown to be very efficient, requiring computations on a very small fraction of the total insulating systems possibilities, 3.8\%, in order to recommend a set of optimized combinations of refractory ceramics materials. The obtained results also allowed further analysis of the optimized set, which highlighted important aspects of furnace designing and insulating product costs and benefits. Additionally, three hyperparameters of the proposed algorithm were optimized such that its search effectiveness was improved while reducing execution time.