Modelagem híbrida URANS-LES para escoamentos turbulentos

• Ano de defesa: 2018
• Autor(a) principal: Elias, Alex José
• 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: Modelagem híbrida URANS-LES; SAS; DES; Escoamentos turbulentos; Cavidade; Degrau; Bluff-body; Prisma; Hybrid URANS-LES; Turbulent flows; Lid-driven cavity; Backward- facing step; Engenharia mecânica; Escoamento turbulento; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA::FENOMENOS DE TRANSPORTE::MECANICA DOS FLUIDOS
• Link de acesso: https://repositorio.ufu.br/handle/123456789/22719; http://dx.doi.org/10.14393/ufu.di.2018.1219

Most of the flows occurring in nature are turbulent. The same can be said about flows of industrial interest. Thus, the differential mathematical modeling of fluid flows enables us to better understand the physical phenomena involved in its occurrence, while providing a large number of valuable information, which are difficult to obtain through material experiments. For this reason, the scientific community has been engaged over the years in developing new modeling strategies for turbulent flows. In this context, the present work presents four hybrid URANS-LES modeling strategies for turbulent flows. DES and SAS hybrid modeling methodologies were implemented in the MFSim code, developed in the Laboratório de Mecânica dos Fluidos of the Universidade Federal de Uberlândia in partnership with PETROBRAS. The implementations take advantage of the URANS and LES basis, characterized by, respectively, the Spalart-Allmaras/k − ω SST models and the Germano dynamic model. Thus, it was possible to implement the SA-DES, SA-SAS, SST-DES and SST-SAS models and test them in the multilevel environment of dynamic adaptive refinement of the MFSim code. It was observed that the hybrid URANS-LES models can be considered an interesting tool in the analysis of turbulent flows with the MFSim code, being able to characterize swirling structures with a behavior closer to the LES methodology than the URANS methodology. In particular, the SA-SAS model presented good results and presents great potential of use within the MFSim code, since it does not have explicit dependence of the computational mesh, like DES models does. The continued use of the SA-SAS model, as well as the sequence of studies about the implementation of other SAS and DES models in the MFSim code, constitutes an interesting strategy in the search for greater robustness in the solution of turbulent flows.