An automotive exhaust system composed of several elements is one
of the most important aggregates in a vehicle in defining the
engine efficiency and passenger comfort. Therefore it is very
essential to understand the flow characteristics inside the exhaust
system thoroughly to keep the back pressure and noise level well
within the limit. The recent widespread use of three-dimensional
(3-D) numerical tools and methods has enabled automotive engineers
to predict the performance of exhaust system at the early stages of
vehicle program.
The present paper describes the application of 3-D computational
fluid dynamics (CFD) using Fluent (V12.0) in an automotive exhaust
system to predict the back pressure and noise concurrently. A 3-D
computational domain consisting of exhaust runner/downpipe,
catalytic-converter (CAT-CON), silencer with internal details and
tail pipe was generated. Extra care was taken by extruding prism
layers while generating the mesh to capture the near wall effects
and smallest acoustic pressure fluctuations. First a steady state
analysis using realizable k-ε turbulence model with non-equilibrium
wall functions was carried out to study the velocity and pressure
field. To resolve the acoustic pressure fluctuation unsteady RANS
simulation was executed followed by large eddy simulation (LES).
Ffowcs-Williams & Hawkings (FW-H) model was used to obtain
frequency versus sound pressure level (SPL) at receiver
location.
The simulation results of flow field, i.e., back pressure and
SPL were compared with experimental results. It confirms that the
exhaust system can be modeled with good accuracy for low back
pressure and improved sound quality using CFD approach with FW-H to
predict the sound pressure level at desired locations. Thus it has
help to reduce the cost and development time by giving better
insights to the design.