Desenvolvimento de técnicas de redução de modelos térmicos para a solução de problemas inversos em condução de calor

• Ano de defesa: 2017
• Autor(a) principal: Ignacio, Luís Henrique da Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Engenharia Mecânica
• Palavras-chave em Português: Engenharia mecânica; Calor; Condução; Problemas inversos (Equações diferenciais); Redução de modelos; Tempo de penetração de calor; Tempo de desvio de calor; Problema inverso; Fluxo de calor; Models reduction; Heat penetration time; Heat deviation time; Inverse problem; Heat flux; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/17989; http://doi.org/10.14393/ufu.di.2016.447

Usually, thermal analysis in practical cases involves transient and three-dimensional models. In these cases, the solution of inverse problems needs complex numerical and analytical solutions, requiring speci c skills to implement and use it. This work is dedicated to the development of techniques to reduce thermal models and solve inverse heat conduction problems. In some cases, tridimensional problems can be reduced to bi or one-dimensional without loss of precision. In others, one-dimensional problems with nite geometry can be reduced to simple geometry problems as the semi-in nite. The solution of the inverse problem in this study aims to estimate di erent kinds of heat ux applied to one-dimensional models. Reduction techniques are based on the concepts of heat penetration and deviation times. Analytical solutions of 1D models, to be studied, are obtained using the Green's functions method. Analysis are applied to samples of Polyvinyl Chloride (PVC) and cemented carbide. The inverse technique applied was TFBGF (Transfer Function Based on Green's Functions), using the impulse response of the semi-in nite model from the heat penetration time until the heat deviation time. Results of the inverse problems, with simulated or experimental data, were satisfactory in heat ux estimate, validating the technique based on heat penetration and deviation times.