Numerical Assessment of Automotive Mufflers Using FEM, Neural Networks, and a Genetic Algorithm

In order to enhance the acoustical performance of a traditional straight-path automobile muffler, multi-chamber muffler having reverse paths is presented. Here, composed two internally parallel/extended tubes and one extended outlet. addition, prevent noise transmission from muffler’s casing, shell also lined with sound absorbing material. Because geometry an automotive complicated, using analytic method predict difficult; therefore, COMSOL, finite element analysis software, adopted estimate loss. However, optimizing shape complicated optimizer linked Finite Element Method (FEM) time-consuming. Therefore, in facilitate optimization, simplified mathematical model used as objective function (or fitness function) during optimization process can be established by Artificial Neural Networks (ANNs) conjunction design parameters related TLs (simulated FEM). With this, proceed linking optimizer, Genetic Algorithm (GA). Consequently, discharged which optimally shaped will improve exhaust noise.