Engine Exhaust noise optimization using Sobol DoE sequence and NSGA-II Algorithms.
Exhaust muffler is the most important component for overall vehicle noise signature. Optimized design of exhaust system plays a vital role in engine performance as well as auditory comfort. Exhaust orifice noise reduction is often contradicted by increased back pressure and packaging space. The process of arriving at exhaust design, which meets packaging space, back pressure and orifice noise requirements, is often manual and time consuming. Therefore, an automated numerical technique is needed for this multi-objective optimization. In current case study, a tractor exhaust system has been subjected to Design of Experiments (DoE) using SOBOL sequencing algorithm and optimized using NSGA-II algorithm. Target design space of the exhaust muffler is identified and modeled considering available packaging constrain. Various exhaust design parameters like; length of internal pipes, location of baffles and perforation etc. are defined as input variables.