Estimativa de fluxo térmico em placas finas empregando termografia por infravermelho e problemas inversos com técnicas de regularização

• Autor(a) principal: Sanches, Emerson de Lima
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UERJ
• Texto Completo: http://www.bdtd.uerj.br/handle/1/13732
• topic: Heat transfer; Inverse problems; Infrared thermography; Regularization techniques; Explicit formulation; Markov Chain Monte Carlo method; Transferência de calor; Técnicas de regularização; Formulação explícita; Calor Transmissão - Modelos matemáticos; Problemas inversos (Equações diferenciais); Termografia; Tecnologia infravermelha; Monte Carlo, Metodo de; Markov, Processos de; CNPQ::CIENCIAS EXATAS E DA TERRA::FISICA::AREAS CLASSICAS DE FENOMENOLOGIA E SUAS APLICACOES::TRANSFERENCIA DE CALOR; PROCESSOS TERMICOS E TERMODINAMICOS

The main objective of the present work is the estimation of thermal fluxes applied to one of the boundaries of a thermally thin sample by solving the associated direct and inverse problems, and by incorporating real temperature measurements obtained experimentally through infrared thermography. For this project, the experimental data obtained through thermography are used for the heat fluxes reconstruction for different test cases. Basically, the experiments are performed for fluxes with variation only in time and with variation only in the two-dimensional space. Thus, the adopted models are, respectively, equivalent to a transient heat conduction problem without spatial variation and to a steady heat conduction problem with spatial variation along two directions. Both models allow the application of an explicit methodology of solution for the inverse problem. This methodology is verified and validated using the Markov Chain Monte Carlo method (MCMC) to solve the same problems, since this method is widely known. In addition, due to the ill-posed nature of the inverse problems, an analysis of the influence of the regularization of the experimental data on the obtained results is carried out. For the explicit methodology, the measured temperatures are replaced by their most important values using truncated eigenfunction expansions. For the MCMC, the Total Variation (TV) technique is used in a priori information. The comparative results between both methods prove the applicability of the explicit method for the addressed problems, as well as demonstrate greater efficiency with regard to the computational-cost compared to the Bayesian and iterative approach via MCMC.