Detalhes do Documento

Modelamento e otimização de filtros acústicos reativos utilizando algoritmos genéticos

Autor(es): Fermiano, Reginaldo Faisca

Origem: OASISbr

Assunto(s): Filtros acusticos; Engenharia mecânica; Otimização matemática; Algoritmos genéticos; Metodo dos elementos finitos


Descrição
Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro Tecnológico, Programa de Pós-Graduação em Engenharia Mecânica, Florianópolis, 2009.Acoustic filters are widely used in the industry to attenuate the noise emission from exhaust machines, cooling ducts, internal combustion engines, inlet and outlet pipes of compressors and others. Reactive acoustic filters are known as mufflers and use a system of tubes and cavities to reflect part of the incident acoustic wave, resulting in a lower transmitted sound pressure than the incident sound pressure. The prediction of acoustic performance of mufflers can be obtained by analytical, experimental and numerical methods as Finite Element Method (FEM) and Boundary Element Method (BEM). Nowadays, in order to obtain more assertive solutions which minimize implementation time and costs, only determine the acoustic performance is not enough. Consequently, optimization methods are important tools to be applied in the industry. This thesis proposes to demonstrate a numerical optimization using the Genetic Algorithm Method (GA) in two reactive mufflers: the first one is assembled with two tubes and one cavity, and the other with three tubes and two cavities. One wide frequency band was chosen for this evaluation. The numerical models were validated using experimental results for the transmission loss, wich was measured using the Two Sources Technique. The results showed FEM that simulations and GA optimization applied with a method of performance prediction in mufflers have good results in build of reactive mufflers in the desirable interest frequencies band with a relative low cost and time of development.
Tipo de Documento Dissertação de Mestrado
Idioma Português
Orientador(es) Lenzi, Arcanjo; Universidade Federal de Santa Catarina
delicious logo  facebook logo  linkedin logo  twitter logo 
degois logo
mendeley logo